HDL Cholesterol

HDL is the "good cholesterol," the number you are supposed to want high. The first half of that reputation holds up: a low HDL genuinely does mark higher cardiovascular risk, and it shows up reliably in study after study. But the second half, the part where raising it makes you healthier, turned out to be one of the most instructive wrong turns in modern cardiology.

For years the logic looked airtight. Low HDL meant more heart attacks, so surely raising HDL would mean fewer. Drugs were built to do exactly that. They raised HDL substantially, and the heart attacks did not budge. That failure, repeated across several large trials, forced a rethink: HDL cholesterol is not a force that protects you so much as a sign of how well the rest of your metabolism is working.

That reframing is what makes this number genuinely useful, as long as you read it correctly. A low HDL is a real signal, usually of insulin resistance and metabolic strain, and worth acting on. But the number itself is a thermometer, not a thermostat. You do not get healthier by forcing it up, and, as it happens, a very high HDL is not better either.

HDL particles carry out reverse cholesterol transport: they pull excess cholesterol from cells, including the cholesterol-laden immune cells in artery walls, and return it to the liver for removal. HDL cholesterol measures the amount of cholesterol those particles are carrying, which is only a rough stand-in for how much of that protective work is actually getting done. The number can be high while the particles function poorly, or adequate while they work well.

This gap between quantity and function is why the simple story broke down. Genetic studies find that people born with naturally higher HDL cholesterol do not have less heart disease, and drug trials that raised HDL with CETP inhibitors and niacin combinations failed to reduce cardiovascular events, which together show that HDL cholesterol mass is not itself causal ^[1]. What appears to matter more is the particles' cholesterol efflux capacity, their actual ability to remove cholesterol, which a standard HDL number does not measure.

So what is a low HDL telling you, if not a deficit of protection? Mostly, it is reporting on the state of your metabolism. Low HDL travels with insulin resistance, excess visceral fat, physical inactivity, and smoking, and it tends to rise when those improve. Read this way, HDL is best understood as a barometer of metabolic health, particularly insulin sensitivity, rather than a shield you can thicken on demand.

None of this means a low HDL can be ignored. As a marker, it carries real information. Even among patients whose LDL had been driven very low with intensive treatment, those with higher HDL still had meaningfully fewer cardiovascular events, so HDL stratifies risk beyond what LDL alone reveals ^[2]. A low number genuinely flags someone who warrants closer attention to the rest of their metabolic and cardiovascular picture.

The mistake is treating the number as a goal. That error runs in both directions, because the relationship between HDL and mortality is U-shaped. In two large Danish cohorts, the lowest mortality sat around 73 mg/dL in men and 93 mg/dL in women, and risk climbed at both extremes; men above roughly 97 mg/dL had about double the all-cause mortality of those at the nadir ^[3]. A strikingly high HDL, often genetic or driven by heavy alcohol use, is not a prize. It frequently signals dysfunctional particles rather than abundant protection.

The practical conclusion is freeing. Because HDL is a marker rather than a target, you do not need to chase it. A low value is a prompt to improve the metabolic health it reflects, which raises HDL naturally as a byproduct, while the actual causal work of lowering cardiovascular risk belongs to the atherogenic particles measured by ApoB and LDL, where the causal evidence is firm ^[4]. Improve the underlying picture and HDL takes care of itself.