Apoprotein B (apo B) is the principle protein that resides in LDL particles along with other proteins, phospholipids, triglycerides, and, of course, cholesterol.
There's a curious thing about apo B. Just like one child per family in China or one television per household in 1950s America, there is only one apo B for every LDL particle.
So measuring apo B, in effect, provides a virtual count of LDL particles. (Actually, VLDL particles, the first lipoprotein to emerge from the liver, also have one apo B per particle but LDL particles far outnumber VLDL particles.) While apo B structure can show limited structural variation from individual to individual, the effect on measured apo B is negligible.
One apo B per LDL particle . . . no more, no less. What about the other components of LDL particles?
The other components of LDL particles are a different story. Cholesterol and triglycerides in LDL particles vary substantially. Diet has profound effects on cholesterol and triglyceride content of LDL particles. A diet rich in carbohydrates, for instance, increases triglycerides in LDL particles while reducing cholesterol. This means that measuring cholesterol in the LDL fraction will be misleading, since cholesterol will be falsely low. LDL cholesterol is therefore a flawed means to assess the behavior and composition of LDL particles. In particular, when LDL particles become enriched in triglycerides, they go through a process that transforms them into small LDL particles, the variety most likely to cause atherosclerosis.
In other words, when the worst situation of all--an abnormal abundance of small LDL particles develops--it is usually not signalled by high LDL cholesterol.
Because apo B is not sensitive to the composition of LDL particles--high cholesterol, low cholesterol, high triglycerides, etc.--it is a superior method to characterize LDL particles. While apo B doesn't tell you whether LDL particles are big, small, or in between, it provides a count of particles that is far more helpful than measuring this deeply flawed thing called "LDL cholesterol."
(Even better: Count LDL particles and measure LDL size, since size gives us insight into sensitivity to oxidation, glycation, adhesiveness, ability to trigger inflammatory pathways via monocyte chemoattractant protein, various interleukins, tunor necrosis factor and others. This is why cholesterol panels should go the way of tie dye shirts and 8-track tapes: They are hopelessly, miserably, and irretrievably inaccurate. Cholesterol panels should be replaced by either apoprotein B or lipoprotein measures.)