Despite recent therapeutic advances, significant residual risk for in-hospital mortality persists among patients admitted with acute myocardial infarction (MI). Low levels of high-density lipoprotein cholesterol (HDL-C), a known independent predictor of increased cardiovascular events, may be an important modulator of heightened risk after acute MI. We evaluated admission HDL-C levels among 98,276 patients with non-ST elevation myocardial infarction with acute MI from the Acute Coronary Treatment and Intervention Outcomes Network Registry–Get With the Guidelines (ACTION Registry–GWTG) program who were enrolled from 490 United States hospitals from January 2007 to December 2010. Clinical characteristics, treatments, atherosclerotic burden, and in-hospital outcomes were analyzed by quartiles of admission HDL-C (Q1: 10 to 30 mg/dl; Q2: 30.1 to 36.9 mg/dl; Q3: 37 to 45 mg/dl; and Q4: 45.1 to 100 mg/dl). Logistic regression was used to explore the relation among HDL-C quartiles, coronary artery disease severity, and in-hospital mortality. Almost half of the patients with acute MI had low admission levels of HDL-C (less than the median 36.9 mg/dl). Such patients were younger, more often men, white, obese, diabetic, smokers, and had higher rates of previous cardiovascular events. After multivariate adjustment, patients with low HDL-C levels had greater extent of severe angiographic multivessel coronary narrowings and higher mortality. Among the 26% of patients in the lowest HDL-C quartile (≤30 mg/dl), there was a 16% greater risk of in-hospital mortality compared with patients in the highest HDL-C quartile (p = 0.012). In conclusion, low levels of HDL-C were common in patients admitted with acute MI and were associated with more extensive angiographic coronary disease. Very low levels of admission HDL-C were observed in one-quarter of patients and associated with significantly higher in-hospital mortality.
In randomized controlled trials of lipid-lowering therapy, significant residual cardiovascular (CV) risk persists despite achievement of optimal levels of low-density lipoprotein cholesterol (LDL-C) with intensive statin therapy. Robust epidemiologic evidence supports a powerful and independent inverse relation between low levels of high-density lipoprotein cholesterol (HDL-C) and incident CV event rates. Normal to high levels of HDL-C mediate several beneficial cardioprotective effects, which include anti-inflammatory, antioxidant, and antithrombotic properties as well as restoration of endothelial homeostasis, whereas conversely, the prothrombotic proinflammatory milieu associated with low levels of HDL-C may be an important pathophysiologic factor in the residual CV risk observed in lipid-lowering trials. Intensive statin therapy early after acute myocardial infarction (MI) reduces recurrent CV events in patients at least in part because of nonlipid-lowering effects. Because HDL-C possesses important pleiotropic properties, it is plausible that low levels of HDL-C levels may adversely impact in-hospital outcomes among patients admitted with acute MI. However, few contemporary studies have evaluated the role of low HDL-C levels in modulating in-hospital outcomes among patients with coronary artery disease (CAD) treated routinely with early intensive statin therapy after acute MI. Accordingly, we used “real-world” observational data from the Acute Coronary Treatment and Intervention Outcomes Network Registry–Get With the Guidelines (ACTION Registry–GWTG) database to evaluate the prevalence of low HDL-C levels on admission, their associated clinical and angiographic characteristics, and the prognostic significance of low HDL-C levels among patients presenting with non-ST segment elevation myocardial infarction (NSTEMI).
Methods
The ACTION Registry–GWTG is a voluntary quality improvement program that includes records from >150,000 existing patients from approximately 600 participating hospitals nationwide. Consecutive patients with a primary diagnosis of ST segment elevation myocardial infarction (STEMI) or NSTEMI are included, as defined by : ischemic symptoms at rest, lasting ≥10 minutes, and occurring within 24 hours of hospital admission, or up to 72 hours for STEMI ; electrocardiographic changes associated with STEMI (new left bundle branch block or persistent ST segment elevation of ≥1 mm in ≥2 contiguous electrocardiographic leads); or positive cardiac biomarkers associated with NSTEMI (creatine kinase-MB or troponin I/T levels that exceed the local laboratory upper limit of normal values) within 24 hours after initial presentation. Participating hospitals collected detailed information on baseline demographic and clinical characteristics, processes of care, and in-hospital outcomes using a standardized set of data elements and definitions. Patients were identified retrospectively through a review of local administrative or clinical databases, and data were entered through a secure, password-protected web-based interface. No informed consent was required as data were anonymous with respect to specific patient characteristics.
We examined records from 260,833 patients enrolled in the ACTION Registry–GWTG from January 1, 2007 through December 31, 2010 at 583 sites. Exclusions included 16,575 patients enrolled with the abbreviated data collection form only (which did not include information on lipid levels), 96,064 subjects with STEMI, 9,165 patients who were transferred out of the index hospital (resulting in inability to track clinical outcomes), 34,244 patients whose HDL-C level was not recorded during the index admission, 395 patients with HDL-C levels <10 mg/dl or >100 mg/dl presumed to be significant outliers, 808 patients with missing HDL-C draw times, and 5,306 patients in whom HDL-C levels were not measured within 48 hours of admission. The final analysis population included 98,276 patients with NSTEMI from 490 sites.
For the concomitant analysis of angiographic characteristics, 15,166 patients who did not undergo diagnostic angiography were excluded. Finally, in computing the adjusted risk of severe CAD (left main or 3-vessel disease vs 1- to 2-vessel disease), 661 patients with missing information on number of narrowed coronary arteries on diagnostic angiography and 6,413 patients with no narrowed epicardial coronary arteries were excluded. Thus, 83,110 patients with NSTEMI from 463 United States hospitals were included in the descriptive analysis of angiographic data, and 76,036 patients from 461 United States hospitals were included in the adjusted severe CAD modeling.
For descriptive analyses, continuous variables are presented as medians with interquartile ranges, whereas categorical variables are expressed as frequencies and percentages. On the basis of the level of HDL-C recorded during hospital admission, subjects were stratified into 4 quartiles (Q1 to Q4). Patients in the lowest quartile (Q1) included patients whose index HDL-C values ranged from 10 to 30 mg/dl; Q2 included patients whose HDL-C values ranged from 30.1 to 36.9 mg/dl; those in Q3 had HDL-C levels from 37 to 45 mg/dl; and Q4 included patients whose HDL-C levels were 45.1 to 100 mg/dl. Clinical characteristics, use of home medications, and hospital treatments and procedures were described across quartiles of HDL-C. Coronary atherosclerotic burden was quantified using mean percent coronary stenoses and the number of diseased coronary arteries. Significant CAD at angiography was defined as a stenosis of ≥50% in the left main coronary artery and/or stenoses of ≥50% in ≥1 of the 3 major epicardial coronary arteries. Using the highest quartile (Q4) as referent, adjusted odds ratios and associated 95% confidence intervals were computed for the risk of severe CAD (left main or 3-vessel disease vs 1- to 2-vessel disease), and risk of in-hospital death for each HDL-C quartile. Logistic generalized estimating equation method with exchangeable working correlation matrix was used for both analyses to account for within-hospital clustering as patients within hospitals tend to have more similar responses relative to patients at other hospitals. This method produces estimates similar to those from ordinary logistic regression, but variances are adjusted for the correlation of outcomes within a hospital. Variables used in the adjusted model for CAD severity included age, gender, race, weight, history of diabetes, hypertension, smoking status, previous MI, previous heart failure, previous percutaneous coronary intervention, previous coronary artery bypass graft surgery, initial serum creatinine, previous stroke, use of home lipid-lowering agent, LDL-C level, and triglyceride level. For the in-hospital mortality analysis, the final adjusted model included variables previously shown to be associated with in-hospital death in the ACTION registry as well as use of home medications and LDL-C and triglyceride levels at presentation. Continuous variables that became nonlinear or flat in relation with CAD severity or in-hospital mortality were fit with linear or truncated splines, respectively. For all covariates included in the models, the percentage of missing data was <3%. Missing variables were imputed to the most frequently occurring group for categorical variables and imputed to the median for continuous variables. For weight and initial serum creatinine, missing values were set to the gender-specific median of the nonmissing values, and all other continuous variables were set the median of the nonmissing values. A p value <0.05 was considered significant for all tests. All analyses were performed using SAS software (version 9.2, SAS Institute, Cary, North Carolina).
Results
The median HDL-C level among 98,276 patients with NSTEMI was 36.9 mg/dl. Clinical characteristics by HDL-C quartiles on admission are listed in Table 1 . Subjects in the lower HDL-C quartiles were younger, more often men, white, had higher body mass index and higher rates of diabetes, and cigarette smoking compared with patients in the higher HDL-C quartiles. The prevalence of previous peripheral artery disease, previous MI, and previous revascularization procedures was likewise greater among patients in lower HDL-C quartiles compared with their counterparts in higher quartiles. Use of aspirin, clopidogrel, β blockers, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker agents, and nonstatin lipid-lowering drugs before hospital admission was greater among patients in lower HDL-C quartiles, consistent with their greater associated co-morbidities. Total cholesterol and LDL-C levels were lowest among patients in the lowest HDL-C quartile (Q1), but total cholesterol/HDL-C ratio, triglycerides, and hemoglobin A1c levels were highest in that group.
| Variable | Overall (n = 98,276) | HDL-C (mg/dl) | |||
|---|---|---|---|---|---|
| 10–30 (n = 25,067) | 30.1–36.9 (n = 23,001) | 37–45 (n = 25,831) | 45.2–100 (n = 24,377) | ||
| Age (yrs) | 65 (55, 76) | 61 (52,72) | 63 (54, 74) | 66 (56, 76) | 70 (59, 81) |
| Men | 62.9 | 77.3 | 71.0 | 61.0 | 42.5 |
| White | 83.2 | 86.7 | 84.5 | 82.3 | 79.3 |
| Body mass index (kg/m 2 ) | 28.6 (25.1, 33.0) | 29.9 (26.4, 34.4) | 29.3 (25.9, 33.5) | 28.4 (25.1, 32.8) | 26.6 (23.3, 30.1) |
| Smoker | 32.5 | 41.4 | 34.7 | 29.2 | 24.6 |
| Hypertension | 74.5 | 74.9 | 74.2 | 74.1 | 74.9 |
| Dyslipidemia | 61.9 | 64.4 | 63.2 | 62.2 | 57.8 |
| Diabetes | 33.1 | 39.1 | 35.3 | 31.6 | 26.5 |
| Previous PAD | 10.8 | 12.0 | 10.7 | 10.3 | 10.3 |
| Previous MI | 26.4 | 30.5 | 27.6 | 25.2 | 22.3 |
| Previous CHF | 13.3 | 14.5 | 12.8 | 12.1 | 14.0 |
| Previous stroke | 8.4 | 8.2 | 8.1 | 8.5 | 9.0 |
| Previous PCI | 23.8 | 27.8 | 25.4 | 23.3 | 18.9 |
| Previous coronary bypass | 17.2 | 20.6 | 18.6 | 16.5 | 13.1 |
| Home medications | |||||
| Aspirin | 46.4 | 47.7 | 46.9 | 46.5 | 44.6 |
| Clopidogrel | 15.3 | 17.0 | 15.8 | 14.8 | 13.8 |
| β Blocker | 40.4 | 43.8 | 42.0 | 39.5 | 36.4 |
| ACEI/ARB | 40.4 | 41.3 | 40.7 | 40.2 | 39.4 |
| Statin | 40.5 | 40.6 | 41.2 | 41.4 | 38.7 |
| Other lipid agent | 10.4 | 12.3 | 11.1 | 9.9 | 8.2 |
| Presenting features | |||||
| ST segment depression or transient ST elevation | 26.2 | 26.3 | 26.5 | 26.7 | 25.4 |
| Signs of CHF | 16.6 | 15.8 | 15.0 | 16.1 | 19.4 |
| Cardiogenic shock | 1.2 | 1.5 | 1.1 | 1.1 | 1.1 |
| Heart rate (beats/min) | 82 (70, 97) | 82 (70, 96) | 81 (69, 96) | 81 (70, 96) | 84 (71, 99) |
| Systolic blood pressure (mm Hg) | 147 (127, 167) | 144 (125, 165) | 147 (128,167) | 148 (129,168) | 148 (128, 169) |
| Initial laboratory data | |||||
| eGFR (ml/min) | 70.6 (53.2, 87.8) | 71.7 (53.3, 89.1) | 71.5 (54.3, 88.6) | 70.5 (53.5, 87.6) | 68.2 (51.6, 86.4) |
| Initial hemoglobin (g/dl) | 13.9 (12.5, 15.1) | 14.1 (12.6, 15.3) | 14.1 (12.7, 15.2) | 13.9 (12.5, 15.0) | 13.4 (12.2, 14.6) |
| Hemoglobin A1C (%) | 6.3 (5.8, 7.5) | 6.5 (5.8, 7.9) | 6.4 (5.8, 7.6) | 6.3 (5.8, 7.5) | 6.1 (5.6, 7.0) |
| Total-C (mg/dl) | 163.0 (135, 196) | 152.0 (123, 185) | 162.0 (133, 194) | 165.0 (137, 198) | 173.0 (147, 203) |
| Total-C/HDL-C ratio | 4.3 (3.4, 5.6) | 5.9 (4.7, 7.2) | 4.8 (3.9, 5.8) | 4.1 (3.4, 4.9) | 3.1 (2.6, 3.8) |
| LDL-C (mg/dl) | 96 (71, 124) | 90 (65, 117) | 98 (73, 125) | 100 (75, 129) | 97 (73, 125) |
| Triglycerides (mg/dl) | 120 (81, 181) | 158 (108, 242) | 135 (94, 198) | 113 (80, 163) | 87 (62, 123) |
Although the use of aspirin within 24 hours was similar across HDL-C quartiles in these patients with NSTEMI, other acute medical therapies and invasive procedures were used more frequently among subjects with lower admission levels of HDL-C ( Table 2 ).
| Variable | Overall, n = 98,276 (%) | HDL-C (mg/dl) | |||
|---|---|---|---|---|---|
| 10–30, n = 25,067 (%) | 30.1–36.9, n = 23,001 (%) | 37–45, n = 25,831 (%) | 45.2–100, n = 24,377 (%) | ||
| Aspirin | 97.7 | 97.7 | 97.9 | 97.8 | 97.4 |
| Clopidogrel | 60.5 | 63.0 | 62.0 | 60.3 | 56.8 |
| β Blocker | 90.8 | 91.3 | 91.4 | 90.8 | 89.5 |
| ACEI/ARB | 49.3 | 50.0 | 50.2 | 49.3 | 47.9 |
| Statin | 62.7 | 63.8 | 64.3 | 63.1 | 59.7 |
| Other lipid agent | 8.1 | 10.3 | 8.9 | 7.5 | 5.9 |
| Any anticoagulant | 93.5 | 93.6 | 93.9 | 93.9 | 92.6 |
| Glycoprotein IIb/IIIa inhibitors | 39.6 | 42.7 | 42.7 | 40.0 | 33.0 |
| Catheterization | 84.6 | 86.3 | 86.9 | 85.2 | 79.9 |
| PCI | 51.6 | 55.5 | 55.1 | 52.0 | 43.9 |
| PCI <48 h | 44.5 | 48.3 | 48.1 | 45.0 | 36.8 |
| CABG | 11.7 | 12.4 | 12.5 | 12.4 | 9.6 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
