18.9 — Cholesterol and Lipid Disorders

Cholesterol is essential. It is in every cell membrane, it is the raw material for every steroid hormone, for vitamin D and for bile acids, and your liver makes most of what you carry (Chapter 1.2).

Nobody has ever had "too little cholesterol" as a clinical problem from diet.

And LDL cholesterol is a cause of atherosclerosis — not a marker, not an association. This has been established by Mendelian randomisation (Chapter 16.1): people who inherit variants giving lifelong lower LDL have proportionally lower cardiovascular disease, which is as close to a randomised trial of lifelong exposure as nature provides.

Both statements are true, and holding them together is the whole of this chapter.

The particles

Cholesterol does not dissolve in blood, so it travels in lipoproteins — particles with a lipid core and a protein shell.

LDL — low-density lipoprotein. Carries cholesterol to tissues, and it is the particle that enters the arterial wall (Chapter 18.1).

HDL — high-density lipoprotein. Carries cholesterol back to the liver.

Triglycerides — carried in chylomicrons from the gut and VLDL from the liver.

Lipoprotein(a) — an LDL particle with an extra protein attached. Almost entirely genetically determined, barely affected by diet or lifestyle, and an independent risk factor.

And it is worth measuring once in a lifetime, particularly with a family history of early cardiovascular disease, because it identifies people at higher risk than their other numbers suggest. Drugs specifically targeting it are in late-stage trials, which would be the first treatment for it.

What the numbers mean now

And the interpretation has changed in three ways.

LDL is the target. The relationship with risk is continuous, log-linear, with no threshold — lower is better, over the range studied, and the benefit depends on both how much and how long.

Non-HDL cholesterol — total minus HDL — is increasingly preferred, because it captures all the atherogenic particles rather than only LDL, and it does not require fasting.

Apolipoprotein B counts the actual number of atherogenic particles, and is arguably better still. Two people with the same LDL can have very different particle numbers, and the one with more, smaller particles is at higher risk.

HDL's role has changed substantially, and this is the honest correction.

HDL remains a good risk marker — low HDL predicts events.

But drugs that raise HDL have consistently failed to reduce events. Niacin, CETP inhibitors, several trials, no benefit.

And Mendelian randomisation finds no causal relationship — genetic variants raising HDL do not lower cardiovascular risk.

So "good cholesterol" is a misleading label. HDL appears to reflect metabolic health rather than to confer protection, and it is no longer a treatment target.

Triglycerides — associated with risk, and the association is partly confounded by the metabolic conditions that raise them. Very high levels — above 10 mmol/L — cause pancreatitis (Chapter 9.5), which is a separate and urgent problem.

Fasting is no longer routinely required for a lipid profile, because non-fasting samples predict risk at least as well.

Diet and cholesterol

And this is where the story has genuinely changed, and where both sides overstate.

Dietary cholesterol has a modest effect on blood cholesterol in most people.

Because the liver makes most of it and downregulates production when intake rises.

Which is why the long-standing advice to limit dietary cholesterol — the eggs and shellfish restriction — was removed from most guidelines. Eggs are not a major driver of blood cholesterol for most people.

Around a quarter of people are "hyper-responders" whose blood cholesterol does respond to dietary intake, which is why the effect is not zero.

Saturated fat raises LDL more than dietary cholesterol does, and the effect is consistent in controlled feeding studies.

And the honest complication: what saturated fat is replaced with determines the outcome. Replacing it with polyunsaturated fat reduces cardiovascular events; replacing it with refined carbohydrate does not.

Which is why "reduce saturated fat" as isolated advice produced the low-fat, high-sugar products of the 1980s and 90s and did not help.

Trans fats are the clearest case — raising LDL and lowering HDL simultaneously, which almost nothing else does (Chapter 1.2). The evidence was strong enough for the WHO to call for global elimination, and many countries have banned or capped them.

Soluble fibre — oats, beans, psyllium — modestly lowers LDL, by binding bile acids in the gut and forcing the liver to make more from cholesterol (Chapter 9.5).

Plant sterols and stanols — added to some spreads — lower LDL by around 10 percent by competing with cholesterol absorption.

And the strongest dietary evidence is for a pattern rather than a nutrient. The Mediterranean diet reduced major cardiovascular events by around 30 percent in the PREDIMED trial — a randomised trial with a hard endpoint, which is rare in nutrition.

Familial hypercholesterolaemia

And this deserves emphasis because it is common, serious, treatable, and massively underdiagnosed.

An inherited defect in LDL clearance — usually the LDL receptor (Chapter 1.4).

Affects around 1 in 250 people — far commoner than most inherited conditions.

LDL is high from birth, so the arterial exposure is lifelong.

Untreated, around half of men have a cardiovascular event by 50 and half of women by 60.

And under 10 percent of people with it are diagnosed in most countries.

Suspect it with: LDL above about 5 mmol/L, a family history of early heart disease, or physical signs — tendon xanthomas, particularly on the Achilles or the knuckles; a white ring around the cornea before 45; and lipid deposits around the eyelids.

Cascade testing of relatives is the highest-yield genetic screening available — each index case identifies several affected family members who can be treated decades before anything happens.

Treatment is statins from a young age, often with ezetimibe and PCSK9 inhibitors, and it normalises life expectancy.

Treatment

Statins — block the rate-limiting enzyme in cholesterol synthesis (Chapter 1.3).

And their benefit is proportional to the LDL reduction, roughly a 20 percent reduction in major vascular events per 1 mmol/L of LDR lowering, sustained over years.

Secondary prevention: NNT around 20 to 40 over five years. Primary prevention in moderate risk: around 100 (Chapter 16.1).

Their side effects, honestly (Chapter 6.7): muscle symptoms occur at almost the same rate on placebo in blinded trials, and rechallenge or switching succeeds in most people who report them. Serious muscle injury is rare. A small increase in new-onset diabetes exists and is substantially outweighed by the cardiovascular benefit.

Ezetimibe — blocks intestinal cholesterol absorption. Adds around 20 percent further LDL reduction to a statin, and reduces events.

PCSK9 inhibitors — injectable antibodies. Very large LDL reductions, for those at highest risk or genuinely statin-intolerant.

Inclisiran — a small interfering RNA reducing PCSK9 production, given twice yearly. A genuinely novel mechanism.

Bempedoic acid — for statin intolerance, acting on the same pathway one step earlier and not activated in muscle.

For high triglycerides: fibrates, high-dose omega-3, and treating the underlying cause — alcohol, uncontrolled diabetes, obesity and some drugs.

Who should be treated

And this is a risk-based decision rather than a number-based one.

Everyone with established cardiovascular disease — secondary prevention, high-intensity statin, regardless of the starting level.

Everyone with familial hypercholesterolaemia.

Most people with diabetes over 40.

People with chronic kidney disease.

And for everyone else, a cardiovascular risk score combining age, sex, smoking, blood pressure, cholesterol and diabetes, with treatment offered above a threshold.

The limitation of risk scores is that age dominates (Chapter 18.1), so a young person with poor risk factors scores low. Lifetime risk and coronary calcium scoring help refine that.

And the decision should be shared. Someone at 12 percent ten-year risk is being offered a treatment that reduces it to around 9 percent — a real benefit, and one people value differently. Presenting it as a mandate rather than a choice is both inaccurate and counterproductive.

What is overstated

In both directions, and it is worth naming.

Overstated by critics: that statins do not work; that cholesterol is irrelevant; that the evidence is fabricated. The evidence base is among the largest in medicine, with hundreds of thousands of participants in randomised trials, and it is consistent.

Overstated by enthusiasm: that everyone over 50 should take a statin regardless of risk; that dietary cholesterol is the main driver; and that a single number determines someone's fate.

And the most useful correction of all: cholesterol is one risk factor among several, and the largest gains for most people come from not smoking, controlling blood pressure, and moving.

What the next page fixes

Chapter 18.10 closes this Part with the clots that form where they should not — deep vein thrombosis and pulmonary embolism, which are common, preventable, and frequently missed.