American Association for Clinical Chemistry
Better health through laboratory medicine
Clinical Laboratory Strategies: October 11, 2007

Evaluating Lipid Measures:
Framingham Researchers Compare Apolipoproteins and Traditional Lipid Ratios for Predicting Coronary Heart Disease
By Julie McDowell


Recent research has indicated that the protein component of LDL cholesterol, apolipoprotein measures or the apo B:apo A-I ratio might be better than traditional lipid measures in predicting coronary heart disease. But many argue that clinicians should stay with the lipid measure most often used now, the total cholesterol:HDL-C ratio. To evaluate this controversy, researchers with Framingham Heart Study compared the performance of different lipid measures, concluding that the overall performance of apo B:apo A-I ratio did not offer significant utility over traditional lipid measures. This issue of Strategies examines these findings and what the results means for the prediction of coronary heart disease.

In the U.S., current guidelines used to detect coronary heart disease (CHD) risk factors primarily target elevated low-density lipoprotein (LDL) cholesterol and total cholesterol. However, research published over the last 10 years has indicated a stronger relationship between high apo B and lower apol A-I levels than with conventional lipids in predicting CHD and other ischemic cardiovascular diseases. For example, results of a Danish study known as The Copenhagen City Heart Study published in March found that apo B was a better predictor of ischemic cardiovascular events compared to LDL cholesterol, because elevated levels were predictive in both men and women (Arterioscler Thromb Vasc Biol 2007;27;661-670).  

This and other studies have suggested that apo B may be a good marker of risk because it is present in several lipoprotein particles, according to Ramachandran S. Vasan, MD, Professor of Medicine, Boston University School of Medicine, and an investigator with the Framingham Heart Study. “Apo B is present in the LDL, very low density lipoproteins, the intermediate density proteins, and the small dense lipoproteins,” he explained. “So in a way, the apo B seems to reflect the total number of atherogenic particles in the blood, which are the particles that are trapped in the arterial wall and cause atherosclerosis.” Because apo B captured all of these particles, many researchers believed that an apo B:apo A-I ratio—the latter representing the major lipoprotein components of high-density lipoprotein (HDL) cholesterol—would be a good risk predictor, he added.

To assess these questions, Vasan and his colleagues analyzed over 3,000 middle-aged participants in the fourth offspring examination cycle (1987-1991) of the landmark Framingham Heart Study, and published their findings in the Aug. 15th issue of JAMA (2777;298[7]:776-785). The researchers evaluated serum total cholesterol, HDL and LDL cholesterol, non–HDL-cholesterol, apo A-I, apoB, and three lipid ratios: total cholesterol [HDL-C, LDL-C]:HDL-C, and apo B:apo A-I. “In our study, we did a head-to-head comparison, looking at how the apo B:apo A-1 ratio measures up to the total cholesterol:HDL-C ratio,” explained Vasan. “Secondly, if you already know the total cholesterol: HDL-C ratio, which is what happens in clinical practice in the U.S., how does knowledge of B:apoA-1 add incrementally to the information we have for predicting risk of coronary disease?”

Their findings showed little difference the total cholesterol and apo ratios, and little additional information gained by the apo measures. “We asked ourselves, if we already know the total cholesterol and HDL-C, does the apo B:apo A-I ratio add to that information in a statistical model that incorporates the former ratio?” said Vasan. “What we found is that if you know the total cholesterol:HDL-C ratio, then the apo A:apo B ratio did not offer incremental information in the same statistical model.”

Evidence to the Contrary

These findings have drawn a strong reaction from many cardiovascular disease researchers. One leading expert, Göran Walldius, MD, Professor of Medicine at the Karolinska Institute in Stockholm and AstraZeneca (Mölndal, Sweden), has been pushing for the inclusion of the apo B:apo A-I ratio in new guidelines for risk evaluation. Walldius was also the lead researcher of the AMORIS (Apolipoprotein-related Mortality Risk) trial, which reported in 2001 that the apo B:apo A-I ratio was of potentially greater value than LDL cholesterol for predicting risk for fatal myocardial infarction after studying over 175,000 individuals including a thousand fatal myocardial events (Lancet, 2001;358[9298]:2026:2033).

Walldius disagrees with the recent JAMA studies conclusion based on a number of reasons, he explained in an e-mail. First, the Framingham cohort of 3,220 subjects is not large enough. “This is especially critical when the predictive power between the total cholesterol:HDL-C and the apo B:apo A-I ratios is compared separately for each gender,” he wrote. “As a comparison, we found in the AMORIS study…that the apo B:apo A-I ratio is clearly the strongest lipid-related risk predictor even compared with the total cholesterol:HDL-C ratio.”

In addition, he pointed out the Framingham population is low risk, meaning that previous CHD patients are excluded and the researchers do not stratify by age. “This should be done, since the predictive power of the apo-ratio is even greater than other lipids and lipid ratios especially in younger subjects who also have the greater risk,” Walldius wrote. “In addition, no comparisons are made between apos, LDL-C and various ratios in patients with normal to low values of LDL-C. Especially in such subjects the apoB:apoA-I ratio is a very good predictor of risk.”

Walldius believes that more clinical research will establish the apo-ratios role in identifying the “at-risk” CHD patients, and therefore improving treatment. “We believe that this new knowledge, once it has been firmly established through even more clinical research, may simplify clinical practice and be more sensitive and specific, since it will better identify and treat the right patient,” he explained. “Therefore, in the longer perspective, this strategy will most likely also be more cost-effective than the older, more conservative LDL-C or even non-HDL-C based approach.”

Was this issue of Strategies of interest to you? Please help us improve by rating this article.