Am Fam Physician. 2014;89(8):662A-662B
Author disclosure: No relevant financial affiliations.
Clinical Question
Is measurement of apolipoproteins better than traditional lipid measurements for predicting cardiovascular risk?
Evidence-Based Answer
Measurement of apolipoprotein B and apolipoprotein A-I is no better than traditional lipid measurements and should not be used to predict cardiovascular risk. (Strength of Recommendation: B, based on meta-analyses with conflicting results.) Apolipoprotein B and non–high-density lipoprotein cholesterol (HDL-C) predict cardiovascular risk slightly better than low-density lipoprotein cholesterol. Elevated levels of apolipoprotein A-I predict a lower risk of cardiovascular events except stroke, but not as well as elevated HDL-C levels.
Evidence Summary
Apolipoproteins are structural components of lipoproteins and have a role in receptor binding and enzyme activation. Apolipoprotein B is carried on all proatherogenic lipoproteins in a 1:1 ratio, and apolipoprotein A-I is found on nearly all HDL particles.1
High levels of apolipoprotein B predict cardiovascular risk about as well as non–HDL-C. A 2012 meta-analysis pooled data from prospective cohort studies of patients without baseline cardiovascular disease and found that non–HDL-C and apolipoprotein B levels were similarly predictive of fatal and nonfatal cardiovascular events2 (Table 12–4 ). Using a clinical model, the authors calculated that substituting total cholesterol and HDL-C measurements with apolipoprotein A-I and B measurements diminished risk prediction by 1% (95% confidence interval [CI], 0.2% to 1.9%), whereas adding them did not significantly improve risk classification. A 2011 meta-analysis of prospective cohort and case-control studies found that apolipoprotein B was a slightly better predictor of cardiovascular risk than non–HDL-C, and both were superior to low-density lipoprotein cholesterol.3
| Number and type of studies | Outcomes measured | Adverse events | Biomarker measured | Risk of elevated biomarker (95% CI) |
|---|---|---|---|---|
| 26 prospective cohort trials (N = 139,581)2 | Fatal and nonfatal coronary artery disease and stroke | 12,234 | Non–HDL-C | HR = 1.27 (1.22 to 1.33)* |
| Apolipoprotein B | HR = 1.24 (1.19 to 1.29)* | |||
| Apolipoprotein A-I | HR = 0.87 (0.84 to 0.90)* | |||
| HDL-C | HR = 0.83 (0.78 to 0.87)* | |||
| 12 (8 prospective cohort and case-control studies; N = 233,455)3 | Fatal and nonfatal ischemic cardiovascular events | 22,950 | Apolipoprotein B | RR = 1.43 (1.35 to 1.51)† |
| Non–HDL-C | RR = 1.34 (1.24 to 1.44)† | |||
| Low-density lipoprotein cholesterol | RR = 1.25 (1.18 to 1.33)† | |||
| 8 randomized controlled trials (N = 38,153)4 | Fatal or nonfatal myocardial infarction, fatal coronary artery disease, hospitalization for unstable angina, and fatal or nonfatal stroke at 1 year | 6,286 | Non–HDL-C | HR = 1.16 (1.12 to 1.19)* |
| Apolipoprotein B | HR = 1.14 (1.11 to 1.18)* | |||
| Low-density lipoprotein cholesterol | HR = 1.13 (1.10 to 1.17)* |
Among patients receiving statins, measurement of apolipoprotein B is comparable to that of non–HDL-C. A 2012 meta-analysis of randomized controlled trials of patients on statin therapy found that time to the first major cardiovascular event was most strongly associated with non–HDL-C levels, followed by apolipoprotein B and low-density lipoprotein cholesterol levels.4 The differences between hazard ratios were small but statistically significant (P = .002 for non–HDL-C vs. low-density lipoprotein cholesterol, and P = .02 for non–HDL-C vs. apolipoprotein B).
Elevated apolipoprotein A-I levels predict coronary events except stroke, but not as well as elevated HDL-C levels. Apolipoprotein A-I levels are inversely associated with cardiovascular disease. In the 2012 meta-analysis discussed previously, comparable inverse associations for cardiovascular risk were seen with HDL-C and apolipoprotein A-I measurements.2 A 2011 prospective cohort study of healthy women found an inverse relationship between apolipoprotein A-I levels and the incidence of stroke and coronary events (myocardial infarction, coronary revascularization, or coronary death).7 Participants included health care professionals 45 years and older who were not receiving lipid-lowering therapy and had a low risk of cardiovascular disease. They were grouped into quintiles based on apolipoprotein A-I or HDL-C levels. The average follow-up was 11.1 years. There were 319 strokes and 602 coronary events among the 26,881 women. Lower apolipoprotein A-I and HDL-C levels were associated with a higher risk of coronary events except stroke. However, in all quintiles the coronary event association with apolipoprotein A-I was weaker than that of HDL-C (hazard ratio for apolipoprotein A-I [lowest vs. highest quintiles] = 1.58 [95% CI, 1.14 to 2.20]; hazard ratio for HDL-C [lowest vs. highest quintiles] = 2.19 [95% CI, 1.51 to 3.19]).7
