LDL is the cholesterol number to act on; HDL is context. "Good" and "bad" cholesterol are nicknames for two delivery vehicles: LDL (low-density lipoprotein) carries cholesterol from the liver out to the body's tissues, and HDL (high-density lipoprotein) brings surplus cholesterol back. The cholesterol inside them is the same molecule. LDL particles build plaque in artery walls and HDL particles do not, which is why guidelines set their treatment goals on LDL.
LDL is the number to act on; HDL is context
- LDL 190 mg/dL (4.9 mmol/L) or above: book a clinician visit and ask about familial hypercholesterolemia.
- LDL 130 to 189 mg/dL: ask for a 10-year heart risk estimate, and whether ApoB and a one-off Lp(a) test would help.
- Low HDL with high triglycerides: ask for HbA1c and fasting glucose to check for insulin resistance.
- Started or changed a cholesterol medicine: recheck in about 4 to 12 weeks 2.
The two markers in detail
Same cholesterol, different carriers
The body needs cholesterol for cell membranes, hormones, bile and vitamin D, but it is a fat, and fat does not dissolve in blood, so the body ships it inside protein-wrapped particles called lipoproteins. The liver sends out VLDL (very-low-density lipoprotein), which sheds its triglycerides and shrinks into LDL. HDL forms separately, picks up spare cholesterol from cells and returns it to the liver. "Density" is the ratio of protein to fat: HDL is mostly protein, so it is the densest; LDL is mostly cholesterol.
The "-C" means cholesterol: a standard lipid panel reports the cholesterol cargo in mg/dL or mmol/L, not the number of particles. An ApoB test counts the plaque-forming particles directly, because each LDL, VLDL and Lp(a) particle (an inherited LDL-like particle) carries one apoB protein; the ApoB vs LDL comparison covers when that count changes the read. Some labs also count HDL particles (HDL-P), but HDL-C is the number guidelines use.
Why LDL is the one that gets treated
LDL particles small enough to slip into the artery wall get trapped there, and the trapped cholesterol builds plaque. The European Atherosclerosis Society reviewed genetic studies, cohort studies and randomised trials together and found that all three point the same way 4. Risk depends on both how high LDL is and how many years it has been high, and across statin trials each 1 mmol/L (about 39 mg/dL) cut in LDL lowered heart attacks and strokes by roughly a fifth 4.
Consistent evidence from numerous and multiple different types of clinical and genetic studies unequivocally establishes that LDL causes ASCVD.Ference et al., European Atherosclerosis Society Consensus Panel, European Heart Journal, 2017 4
ASCVD is atherosclerotic cardiovascular disease: heart attacks and strokes caused by plaque. That is why the 2018 AHA/ACC guideline builds its treatment decisions around LDL, with non-HDL cholesterol and ApoB as secondary measures 2. For an adult aged 20 to 75 with an LDL of 190 mg/dL (4.9 mmol/L) or higher, the guideline recommends that clinicians consider a high-intensity statin, a daily cholesterol-lowering medicine, without first running a risk calculation 2; whether and how to treat is a decision to make with your clinician. A level that high also raises the question of familial hypercholesterolemia, an inherited cause of very high LDL.
Why "raise your HDL" stopped being the advice
In large cohort studies, people with higher HDL have fewer heart attacks, which is where the "good cholesterol" label came from 5. The trouble started when researchers tried to raise HDL on purpose. People born with a gene variant that lifts HDL cholesterol for life had no lower heart attack risk than people without it 5. Adding niacin to statins raised HDL in two large trials, AIM-HIGH and HPS2-THRIVE, and did not cut heart attacks or strokes 2. The drug torcetrapib raised HDL by about 72% and increased deaths 7.
These data challenge the concept that raising of plasma HDL cholesterol will uniformly translate into reductions in risk of myocardial infarction.
The ESC/EAS guideline now states plainly that HDL cholesterol is not a treatment target 3. Very high HDL also stops looking protective. In about 116,000 Danish adults, men with HDL of 97 mg/dL (2.5 mmol/L) or more and women with 116 mg/dL (3.0 mmol/L) or more had higher all-cause mortality than people in the middle 6.
Low HDL still earns its place on the panel. An HDL under 40 mg/dL in men or under 50 mg/dL in women, especially with high triglycerides, is one of the defining features of metabolic syndrome and often the first lipid sign of insulin resistance, when the body needs more insulin than normal to keep blood sugar steady. The page on low HDL walks through the usual causes.
Normal ranges for LDL and HDL
Where an example LDL of 132 mg/dL fallsmg/dL
| Marker | Band | mg/dL | mmol/L |
|---|---|---|---|
| LDL | Optimal | Under 100 | Under 2.6 |
| LDL | Near optimal | 100 to 129 | 2.6 to 3.3 |
| LDL | Borderline high | 130 to 159 | 3.4 to 4.0 |
| LDL | High | 160 to 189 | 4.1 to 4.8 |
| LDL | Very high | 190 or more | 4.9 or more |
| HDL | Low, men | Under 40 | Under 1.0 |
| HDL | Low, women | Under 50 | Under 1.3 |
| HDL | Counted as protective in older guidance | 60 or more | 1.6 or more |
These bands come from the US NCEP Adult Treatment Panel III 1. The newer ESC/EAS guideline sets LDL goals by risk instead: under 116 mg/dL (3.0 mmol/L) at low risk, under 100 (2.6) at moderate risk, and under 70 (1.8) at high risk and under 55 (1.4) at very high risk, with the high and very high risk goals also asking for at least a 50% cut from the starting level 3. Lab reference ranges vary, and an LDL inside the printed range is not automatically your target. For single values, see what an LDL of 100 or an LDL of 120 means, and the lipid panel guide for the whole report.
CalculatorCholesterol converterYour report in mmol/L and the cutoffs in mg/dL? Convert either way.Work it out Reading LDL and HDL together: four common patterns
| Pattern | What it usually means | What to look at next |
|---|---|---|
| High LDL, low HDL, high triglycerides | The classic higher-risk profile: more plaque-forming particles plus signs of insulin resistance | ApoB, glucose and HbA1c, blood pressure |
| High LDL, high HDL | The high HDL does not cancel the LDL; guidelines judge the LDL against overall risk, not against HDL | LDL or ApoB, family history, Lp(a) |
| Normal LDL, low HDL, high triglycerides | The insulin-resistance pattern; LDL-C can understate how many LDL particles are present | ApoB, non-HDL cholesterol, triglyceride to HDL ratio |
| Very high HDL, everything else normal | Can be genetic or linked to alcohol intake; adds no protection, and at the extremes (97 mg/dL or more in men, 116 or more in women) is linked to higher mortality, so worth a clinician's look | Alcohol intake, the rest of the panel |
Markers that settle the pattern
Very high triglycerides change the read in their own right. A triglyceride level above 500 mg/dL raises the risk of pancreatitis 2, so severe upper belly pain with very high triglycerides needs urgent care; the page on high triglycerides covers the bands, and the high ApoB page covers what a high particle count adds.
Ratios, non-HDL and ApoB: the better summaries
Neither the AHA/ACC nor the ESC/EAS guideline sets a treatment goal on the LDL to HDL ratio 2 3. Non-HDL cholesterol (total cholesterol minus HDL) captures the cholesterol in every plaque-forming particle in one number and needs no extra test. ApoB counts those particles, and the AHA/ACC guideline counts an ApoB of 130 mg/dL or more as a risk enhancer, a finding that tips a borderline decision towards treatment 2. The ESC/EAS guideline also recommends measuring Lp(a) at least once in a lifetime 3; see high Lp(a).
Your cholesterol ratios, in context
Two ratios that say more than any single number. Nothing leaves your browser.

Risk calculators still use HDL. The 10-year heart risk score UK GPs use takes the total cholesterol to HDL ratio, and US risk equations take total cholesterol and HDL as separate inputs, which is why the total cholesterol to HDL ratio and the triglyceride to HDL ratio appear on many reports. The cholesterol ratio calculator works out all three from a standard panel.
How the test works: fasting, calculated LDL, how often
Fasting is usually unnecessary: both the AHA/ACC and ESC/EAS guidelines accept a non-fasting sample for routine testing 2 3, with a fasting repeat if triglycerides come back at 400 mg/dL or more 2. For adults aged 20 and over, NCEP ATP III suggests a lipid panel about every 5 years, more often with risk factors or on treatment 1.
How to lower LDL and raise HDL
LDL falls most with less saturated fat, more soluble fibre (oats, beans, lentils) and weight loss, and with medication when overall risk warrants it; the foods that lower cholesterol page lists what the trials support. If LDL stays high on a good diet, the cause is often inherited, but an underactive thyroid, kidney or liver disease, or some medicines can also raise it, so this is a conversation for your clinician 2. The high cholesterol page covers the medical options.
HDL rises modestly with regular aerobic exercise, stopping smoking, losing excess weight and cutting refined carbohydrates 3. Those habits lower heart risk through blood pressure, blood sugar and weight, and a few extra points of HDL come along as a by-product. After starting or changing a cholesterol medicine, guidelines suggest rechecking the panel in about 4 to 12 weeks 2.
Questions
Which is more important, HDL or LDL?
LDL. LDL particles cause plaque, and lowering LDL lowers heart attacks and strokes in trials, so guidelines treat LDL. HDL is a risk clue: a low HDL is a warning flag, but raising HDL has not been shown to prevent heart disease.
Can a high HDL cancel out a high LDL?
No. People with genes that raise HDL had no fewer heart attacks, and drugs that raised HDL did not prevent them. Guidelines judge a high LDL against overall risk, and a high HDL does not lower that; your clinician weighs both.
Can HDL be too high?
Yes, at the extreme. In Danish data, HDL of 97 mg/dL or more in men and 116 mg/dL or more in women was linked to higher all-cause mortality. Moderately high HDL is usually fine, though it adds no extra protection.
Can LDL be too low?
Very low LDL reached on treatment has not caused added problems in trials so far, according to the ESC/EAS guideline. A very low calculated LDL can be a formula error, so a direct LDL or ApoB checks it. A low LDL without treatment is worth mentioning to your clinician.
What is a good total cholesterol to HDL ratio?
Lower is better. Many labs flag a ratio above 5, but neither major guideline sets a cutoff, because the ratio is an input to risk scores rather than a target. The cholesterol ratio calculator works it out from your panel.
Do I need to fast for a cholesterol test?
Usually not. US and European guidelines accept non-fasting samples for routine lipid testing. A fasting repeat helps if triglycerides come back at 400 mg/dL or more.
Why is my LDL high when I eat well?
LDL is strongly inherited. An underactive thyroid, kidney or liver disease and some medicines can also push it up, so a clinician checks for these. An LDL of 190 mg/dL or more, or a family history of early heart disease, raises the question of familial hypercholesterolemia, which needs a clinician's assessment, since diet alone often is not enough.
Citations
- National Cholesterol Education Program Expert Panel. Third Report of the NCEP Adult Treatment Panel III, Final Report. Circulation. 2002;106(25):3143-3421.
- Grundy SM, Stone NJ, et al. 2018 AHA/ACC Guideline on the Management of Blood Cholesterol. Circulation. 2019;139(25):e1082-e1143.
- Mach F, Baigent C, et al. 2019 ESC/EAS Guidelines for the management of dyslipidaemias. European Heart Journal. 2020;41(1):111-188.
- Ference BA, Ginsberg HN, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. EAS Consensus Panel. European Heart Journal. 2017;38(32):2459-2472.
- Voight BF, Peloso GM, et al. Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study. The Lancet. 2012;380(9841):572-580.
- Madsen CM, Varbo A, Nordestgaard BG. Extreme high high-density lipoprotein cholesterol is paradoxically associated with high mortality in men and women. European Heart Journal. 2017;38(32):2478-2486.
- Barter PJ, Caulfield M, et al. Effects of torcetrapib in patients at high risk for coronary events (ILLUMINATE). New England Journal of Medicine. 2007;357(21):2109-2122.
Educational context only, not medical advice or a diagnosis. Always discuss your results with a clinician.