For decades, researchers have sought to understand the cardiovascular benefits of high-density lipoprotein (HDL) cholesterol, long considered the “good” kind of cholesterol for its ability to transport low-density lipoprotein (LDL) cholesterol out of the blood stream. Yet clinical trials of drugs designed to improve HDL levels and transport have not demonstrated cardiovascular benefit.

Now a new study by Israeli researchers, published in the journal Atherosclerosis, provides a clue to the conundrum: HDL’s benefits may rest with the levels of subcategories of HDL particles, rather than with the total HDL-C level.

HDL-C is a complex, heterogeneous form of cholesterol with differences based on size, density, charge, chemical composition and functionality. For instance, smaller, denser, protein-rich HDL particles (HDL3) are more atheroprotective than large, buoyant, cholesterol-rich particles (HDL2). In addition, HDL particle concentration (HDL-P), which measures the total number of particles into which HDL cholesterol is packaged, is another predictor of coronary heart disease (CHD). In fact, its inverse relationship with CHD is strong enough that it may one day replace total HDL-C as a CHD predictor.

Researchers from Hebrew University-Hadassah School of Public Health and Community Medicine in Jerusalem used nuclear magnetic resonance (NMR) spectroscopy to identify the numbers and sizes of plasma HDL particles in participants’ blood, and helical CT-scanning to identify coronary artery calcification (CAC). They then looked for associations between the concentrations and sizes of different HDL particles and CAC.

The investigators’ goal was to see if the newer findings for the HDL2 and HDL3-subclasses, and for HDL-P, persist across different populations, laboratory methods, and outcome measures.

The researchers found statistically significant multivariable-adjusted inverse associations of medium-sized (MS) HDL-P and HDL-P with CAC scores >100 in participants without diabetes who were not on lipid-lowering medications. They found no association between large HDL-P and average HDL-P and CAC, and a protective association of HDL-C with CAC only in men. There was no statistically significant interaction of ethnicity or sex with HDL-P or MS-HDL-P.

Thus, they concluded, while HDL-C is an “indisputable risk marker for CHD, HDL-P and MS-HDL-P are better independent markers of CAC.”

“Restricting the assessment of HDL to HDL-C measurement alone, which is more reflective of the concentration of large HDL particles, suggests that HDL-C may be a poor surrogate for HDL function,” they wrote. “With a better current understanding of the complexity of the HDL particle and its properties, and the ability to measure the HDL moieties, it is now possible to move past HDL-C to more refined measures of HDL that appear to better reflect functional roles of HDL and CHD risk assessment.” Such tests, they concluded, should be considered as part of the risk assessment for CHD.