📚 Part of our Heart Health Guide — an evidence-based guide.
For decades the advice on LDL cholesterol was a single number that applied to almost everyone: get under 100, and under 70 if you have heart disease. Then the targets kept moving down. Guidelines now discuss 55. Some cardiologists argue for 40. Trials have driven patients to a median LDL of 30 and reported no harm.
That drift makes people suspicious, and reasonably so. Either the earlier numbers were wrong, or something is being oversold. The answer is more interesting than either — and the honest version includes a limitation that rarely makes it into the headline.
Why LDL is causal, not just correlated
The strongest argument for lowering LDL does not come from drug trials at all. It comes from genetics.
Some people are born with variants — in PCSK9, HMGCR, NPC1L1 — that give them modestly lower LDL for their entire lives. Studies comparing these people to the general population (a design called Mendelian randomisation) consistently find substantially lower rates of coronary disease.
This design is powerful because gene variants are allocated at conception, essentially at random, and cannot be confounded by diet, exercise or income the way observational studies are.
What it showed is striking: a lifetime of slightly lower LDL protects far more than a large reduction started at 60. Roughly three times more benefit per unit of LDL lowered, in several analyses. LDL exposure appears to be cumulative — it is the area under the curve across decades, not your number this year.
That single finding reframes the entire question. It means the real variable is how long, not only how low.
What the drug trials found
The Cholesterol Treatment Trialists’ Collaboration pooled data from many randomised statin trials and found a consistent relationship: each 1 mmol/L (about 39 mg/dL) reduction in LDL produced roughly a 22% reduction in major vascular events per year of treatment, with the benefit accruing over time.
The question that remained was whether this held at very low levels, or whether the benefit flattened out.
IMPROVE-IT added ezetimibe to simvastatin in patients after acute coronary syndrome, driving LDL to about 54 mg/dL against 70 in the comparison group. The lower group had fewer events. Modest difference, but it demonstrated the principle held below 70, and with a non-statin drug — suggesting the benefit comes from the LDL reduction itself, not from something unique to statins.
FOURIER added evolocumab, a PCSK9 inhibitor, to statin therapy and reached a median LDL of 30 mg/dL. Events fell further. Analyses of participants who reached the very lowest levels — below 20 mg/dL — found no increase in adverse outcomes.
Across these trials, no lower threshold at which LDL lowering starts causing harm has been identified. Concerns raised over the years about very low LDL and cancer, haemorrhagic stroke or cognitive impairment have not been borne out. A dedicated cognitive study, EBBINGHAUS, found no decline.
The limitation that gets left out
Here is what an honest reading requires.
Relative risk reduction is not absolute benefit. A 22% reduction sounds transformative. What it means for you depends entirely on your starting risk. For someone who has already had a heart attack, that translates into a meaningful number of prevented events. For a 45-year-old with no risk factors and a 2% ten-year risk, the same relative reduction moves an already small number slightly.
This is why blanket targets mislead. The aggressive numbers were established in high-risk populations — people with established cardiovascular disease, familial hypercholesterolemia, or diabetes with additional risk factors. Extending them to a healthy 40-year-old with an LDL of 130 is an extrapolation, not a finding.
Also worth stating: most trials of very low LDL ran for two to three years. Long-term safety at an LDL of 20 over three decades has not been directly observed. The genetic evidence is reassuring on this point — people with lifelong low LDL do well — but it is inference, not a completed trial.
The targets, and who they are for
| Risk category | Common LDL target |
|---|---|
| Very high risk — prior heart attack, stroke, or revascularisation | Below 55 mg/dL, and at least a 50% reduction |
| High risk — diabetes with organ damage, familial hypercholesterolemia, calcium score above 100 | Below 70 mg/dL |
| Moderate risk | Below 100 mg/dL |
| Low risk, no other factors | No target; focus on lifestyle and reassessment |
European guidelines are more aggressive than American ones on the lowest tier — a genuine disagreement between expert bodies reading the same evidence, and a reminder that these are judgement calls about risk-benefit, not measurements of a natural constant.
When the number alone is not enough
Two things change the interpretation of an LDL result, and both are underused.
Lipoprotein(a). Inherited, unaffected by lifestyle, missing from a standard panel, and elevated in roughly one in five people. It is a reason to treat LDL more aggressively, since you cannot lower Lp(a) itself (why one test lasts a lifetime).
A coronary calcium score. This measures whether atherosclerosis is actually present rather than estimating whether it might be. It frequently changes the decision in the intermediate-risk group, in both directions (how to read one).
Also relevant: high triglycerides with low HDL points to insulin resistance driving the picture, which needs a different response than LDL alone (what triglycerides mean, and how to reverse insulin resistance).
What lowers LDL without a prescription
Diet moves LDL meaningfully but not unlimitedly — typically 10 to 15%, occasionally more. That is genuinely useful and it is not equivalent to what a statin does.
What works: replacing saturated fat with unsaturated fat rather than with refined carbohydrate; viscous soluble fiber, 5–10 g daily from oats, barley, psyllium and legumes (how it works); plant sterols; nuts, which have randomised evidence behind them (what PREDIMED found); and the overall Mediterranean pattern (the evidence). Exercise does more for triglycerides and HDL than for LDL, which surprises people.
For the full non-drug approach, see how to lower cholesterol naturally, and what each number on the panel means.
When to see a doctor
Ask for a lipid panel if you have not had one in five years, or sooner with a family history of early heart disease. See a doctor promptly if your LDL is above 190 mg/dL, which raises the possibility of familial hypercholesterolemia — a condition affecting roughly 1 in 250 people, largely undiagnosed, where treatment decisions differ and family members need testing.
Discuss statin side effects rather than stopping treatment on your own; most reported symptoms are not caused by the drug and there are alternatives (what the blinded trials found).
Seek emergency care for chest pressure, pain radiating to the arm or jaw, sudden breathlessness, or one-sided weakness — regardless of any cholesterol number.
The honest bottom line
Lower LDL is better, the relationship holds down to levels far below old targets, and no threshold of harm has been found. The strongest evidence is genetic and shows that duration of exposure matters more than any single reading — which argues for acting earlier rather than harder. But the aggressive targets were built in high-risk populations, and applying them to someone at low risk converts a real finding into an overstatement. Your LDL number means little without your risk context, and a calcium score or an Lp(a) often tells you more than another decimal place.
Frequently asked questions
What is a good LDL cholesterol level?
It depends entirely on your cardiovascular risk. For someone with prior heart attack or stroke, guidelines commonly target below 55 mg/dL. For high risk such as familial hypercholesterolemia or a calcium score above 100, below 70. For moderate risk, below 100. For a low-risk person with no other factors, there is no specific target and the focus is lifestyle plus periodic reassessment. A single LDL number means little without the risk context around it.
Can LDL cholesterol be too low?
No lower threshold at which LDL lowering begins to cause harm has been identified in trials. FOURIER drove median LDL to about 30 mg/dL, and participants reaching below 20 showed no increase in adverse events. Earlier concerns about cancer, haemorrhagic stroke and cognitive decline have not been borne out — the EBBINGHAUS study specifically found no cognitive decline. The honest caveat is that these trials ran two to three years, so multi-decade safety at very low levels is inferred from genetic evidence rather than directly observed.
How much can diet lower LDL cholesterol?
Typically 10 to 15%, occasionally more. The most effective changes are replacing saturated fat with unsaturated fat rather than with refined carbohydrate, eating 5 to 10 grams daily of viscous soluble fiber from oats, barley, psyllium or legumes, adding plant sterols, and eating about an ounce of nuts most days. That is a genuine and worthwhile reduction, but it is not equivalent to what statin therapy achieves, and framing it as a substitute in high-risk patients is misleading.
Does a high LDL always mean I need a statin?
No. The decision depends on your absolute risk, not the LDL alone. An LDL of 130 in a low-risk 40-year-old is a different situation from the same number in someone with diabetes or a strong family history. A coronary calcium score often resolves the question in the intermediate-risk group by showing whether atherosclerosis is actually present, and it can move the decision in either direction. An LDL above 190, however, warrants prompt medical attention regardless.
