Biology › Nutrition, digestion and health evidence › Cholesterol, atheroma and what a risk factor licenses

Cholesterol, atheroma and what a risk factor licenses

Your liver makes more cholesterol before breakfast than most breakfasts contain, because without it you would have no membranes, no steroid hormones, no bile salts and no vitamin D. The problem is never the molecule. It is what happens when the particles carrying it get into an artery wall and stay there.

Before this Cholesterol in the phospholipid bilayer · Receptor-mediated endocytosis · The structure of arteries and the cardiac cycle · Correlation, confounding and study design

What you should be able to do

• Describe cholesterol as a sterol and state four things the body needs it for.
• Explain what a lipoprotein is and distinguish LDL from HDL by composition and function.
• Describe the formation of an atheroma in sequence, from endothelial damage to a fibrous plaque.
• Explain how a plaque leads to angina, to a myocardial infarction and to an aneurysm.
• State precisely what the term 'risk factor' licenses you to claim, and what it does not.
• Account for why the evidence on LDL is regarded as causal while the evidence on HDL is not.

Cholesterol is not a fat, and it is not optional

Cholesterol is a lipid but not a triglyceride. It is a sterol: four carbon rings fused together, with a short hydrocarbon tail at one end and a single hydroxyl group at the other. That hydroxyl is the only polar part of an otherwise hydrophobic molecule, which is exactly the arrangement needed to sit in a membrane with its hydroxyl among the phosphate heads and its rings among the fatty acid tails.

Four jobs are worth naming. Cholesterol regulates the fluidity of every cell surface membrane you have, restricting movement of the phospholipids at higher temperatures and preventing them packing too closely at lower ones. It is the precursor of the steroid hormones — testosterone, oestrogen, cortisol, aldosterone — and of the bile salts that emulsified the fat in the previous lesson. And it is the starting material for vitamin D.

Most of it is not eaten. The liver synthesises on the order of a gram a day, while a typical diet supplies two or three hundred milligrams, and the liver adjusts its synthesis downwards when dietary intake rises. The rate-limiting enzyme of that synthetic pathway is HMG-CoA reductase, and inhibiting it is what a statin does — which is why statins lower blood cholesterol far more effectively than removing cholesterol from the diet does.

That regulation is why the long-standing advice to limit dietary cholesterol was dropped from several national guidelines around 2015: in most people, eating more of it produces a small rise in blood cholesterol and a compensating fall in synthesis. Eggs turned out to be a poor lever on the quantity everyone was trying to move.

Lipoproteins: the vehicle, not the cargo

Cholesterol and triglyceride are insoluble in plasma, which is mostly water, so they cannot simply dissolve and travel. They are carried inside particles built for the purpose.

A lipoprotein has a core of cholesteryl esters and triglyceride, wrapped in a single layer of phospholipid with the hydrophilic heads facing outwards, with free cholesterol among them and one or more apoproteins threaded through the surface. The apoprotein is the address label: it determines which receptors the particle can bind to and therefore where it is unloaded.

LDL is the delivery service. A cell needing cholesterol makes LDL receptors and places them in its surface membrane; an LDL particle binds and the whole particle is taken in by receptor-mediated endocytosis and digested in a lysosome. A cell with plenty of cholesterol makes fewer receptors, so uptake is regulated by need, and LDL not taken up stays in the plasma.

HDL runs the other way. Nascent HDL particles are mostly protein and accept cholesterol from cell membranes, including from cells inside artery walls, and carry it back to the liver, which can excrete it in bile. That is reverse cholesterol transport, and it is where HDL's reputation came from.

The examinable comparison is not 'good and bad'. It is that LDL particles are larger, cholesterol-rich and delivering, while HDL particles are smaller, protein-rich and collecting, and that a high LDL concentration means more particles available to enter an artery wall.

How an atheroma is built

Atherosclerosis takes decades and happens in a fixed order. Fatty streaks are present in the aortas of many teenagers; the first symptom typically arrives in the sixth decade or later. Nothing in this sequence happens quickly except the last step, which happens in minutes.

It starts with damage to the endothelium, the single layer of cells lining the vessel. High blood pressure provides mechanical stress, particularly where arteries branch; chemicals in tobacco smoke, high blood glucose and a high LDL concentration all injure the lining as well. An intact endothelium is smooth and does not admit much.

Next, LDL particles pass into the wall beneath the endothelium and are retained there, where they are chemically modified — oxidised. Modified LDL is treated as damage rather than as cargo. White blood cells, monocytes, adhere to the activated endothelium and migrate into the wall, becoming macrophages, and take up the modified LDL through receptors that are not regulated by how much cholesterol the cell already holds. So they keep eating, fill with lipid, and become foam cells. A collection of foam cells is a fatty streak, and this is an inflammatory process from beginning to end.

Smooth muscle cells then migrate from the deeper layers of the artery wall, multiply, and deposit collagen over the accumulating lipid, forming a fibrous cap. The result is an atheroma or plaque: a lipid and cellular core under a fibrous cap, bulging into the lumen and reducing the space for blood to flow.

Atheroma

A plaque in the wall of an artery, consisting of lipid, foam cells and cell debris beneath a fibrous cap, which narrows the lumen and stiffens the wall.

Thrombosis

The formation of a blood clot inside a blood vessel, most often on a ruptured atheroma, which may block the vessel completely.

Aneurysm

A localised bulge in an artery wall that has been weakened, which may rupture and cause severe internal bleeding.

Two consequences follow, one gradual and one sudden. Gradually, a narrowed coronary artery delivers too little blood when demand rises, so the muscle it supplies runs short of oxygen during exertion and the person feels the pain of angina, which stops on resting.

Suddenly, the fibrous cap tears. Collagen beneath it is exposed to the blood, platelets adhere and release thromboplastin, thromboplastin with calcium ions catalyses the conversion of prothrombin to thrombin, and thrombin catalyses the conversion of soluble fibrinogen into insoluble fibrin, which forms a mesh trapping cells. That clot can occlude the artery within minutes. In a coronary artery the muscle downstream is deprived of oxygen, cannot respire aerobically, and dies: a myocardial infarction. In an artery supplying the brain the result is a stroke. A fragment breaking away and lodging elsewhere is an embolus.

A plaque also weakens the wall it sits in. An artery carries blood at high pressure, and a weakened section stretches, which thins it further and lets it stretch more: an aneurysm. Rupture of a large one, in the abdominal aorta, is usually fatal, which is why the NHS invites men to a single ultrasound scan in the year they turn 65 — the age and sex at which the condition is common enough for screening to find more than it harms.

Risk factors, and what the phrase permits you to say

A risk factor is a variable associated with an increased probability of a disease in a population. That is the whole definition, and it is worth being precise about what follows from it and what does not.

It licenses prediction. Knowing someone's age, sex, blood pressure, smoking status, cholesterol concentration and postcode allows a reasonable estimate of their probability of a cardiovascular event in the next ten years, and the tools used in general practice do exactly that. In the UK a ten-year risk above 10% is the point at which a statin is normally offered and discussed.

It does not license three things. It does not establish a mechanism — an association can exist because the factor causes the disease, because something else causes both, or because early disease causes the factor. It does not predict an individual: a risk of 12% means about twelve of a hundred similar people will have an event, and says nothing about which twelve. And it does not promise that removing the factor removes the risk, which is a separate question needing a trial or something as good as one.

A case-control study published in 2004, covering around 15,000 first heart attacks in 52 countries, found that nine measurable factors accounted for the great majority of the risk of a first myocardial infarction in every region studied — smoking, blood lipids, blood pressure, diabetes, abdominal obesity, psychosocial factors, fruit and vegetable intake, alcohol and physical activity. Notice two things together: consistency across very different populations is powerful evidence, and it was still a case-control study, with everything that implies about recall and about choosing controls.

How the evidence was gathered, and what is still argued about

The claim that LDL causes atherosclerosis is about as secure as any causal claim in medicine, and the way it was established is worth knowing, because it is the model for what strong evidence looks like when a randomised trial of a lifetime exposure is impossible.

The first strand is a natural experiment. Familial hypercholesterolaemia is caused by a fault in the gene for the LDL receptor, so cells take up LDL poorly and the blood concentration is roughly doubled from birth. It affects around 1 in 250 people and, untreated, produces heart attacks decades early. The raised LDL cannot be a consequence of an unhealthy life, because it was set at conception.

The second strand runs the same logic the other way. Some people carry inactive versions of a gene called PCSK9 and have LDL concentrations around a quarter lower than average for their whole lives. Their rate of coronary heart disease is lower by far more than a quarter. Because the genetic variant is allocated at conception and is not associated with income, smoking or diet, this kind of analysis behaves like a randomised trial run by nature, and it is called Mendelian randomisation.

The third strand is intervention. Trials of drugs that lower LDL by unrelated mechanisms — statins, ezetimibe, PCSK9 inhibitors — lower cardiovascular events in proportion to how much LDL they lower. Genetics, mechanism and trials agree, which is the pattern that makes a causal claim safe.

Now the contested part, and it is instructive precisely because the same field produced both. HDL concentration predicts cardiovascular risk reliably: people with more of it have fewer events. But raising it has repeatedly failed to help. Drugs that raised HDL substantially did not reduce events, and one had to be abandoned when the treated group did worse. Mendelian randomisation tells the same story: genetic variants that raise HDL are not associated with lower heart disease, while variants that raise LDL are associated with more of it. The honest summary is that HDL is a marker of something protective rather than established as the protective thing itself, and 'good cholesterol' was always a slogan rather than a finding.

Dietary fat is contested at a different level. Controlled feeding studies, in which every meal is prepared and measured, showed decades ago that replacing saturated fat with polyunsaturated fat lowers blood LDL, and that part is not seriously disputed. What happens to heart attacks and deaths is less settled: a systematic review of the randomised trials found that reducing saturated fat intake cut combined cardiovascular events by about 17%, with no clear effect on death from any cause. What it is replaced with matters as much — unsaturated fat looks beneficial, refined carbohydrate does not.

Two episodes show how an evidence base itself can be distorted. Data from two large controlled feeding trials run in the 1960s and 1970s were only fully published after being recovered from the investigators' records forty years later, and showed that lowering blood cholesterol by changing dietary fat did not translate into fewer deaths in those trials. And documents recovered in 2016 showed that a 1967 review of sugar and heart disease in a leading journal had been funded, undisclosed, by the sugar industry, which chose the papers discussed. Neither overturns the biology. Both are reasons to ask who funded a study, whether it was registered before it began, and whether the result would have been published had it come out the other way.

In the exam

• Never write 'good cholesterol' or 'bad cholesterol' in an answer. Write 'cholesterol carried in HDL' and 'cholesterol carried in LDL', and the biology behind the sentence becomes visible to the examiner.
• Atheroma questions want the sequence in order: endothelial damage, LDL entering and being modified, macrophages becoming foam cells, smooth muscle and collagen forming a cap, narrowing of the lumen. Marks are usually allocated per step.
• For a heart attack, the required chain is: cap ruptures, platelets adhere, clot forms, coronary artery blocked, cardiac muscle deprived of oxygen, respiration and therefore ATP supply fails, muscle dies.
• 'Risk factor' means associated with increased probability in a population. If a question asks whether a factor causes the disease, answer the question actually asked and say what further evidence would be needed.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

← Diet, energy and how to read a health claim · Pathogens, and the defences that never ask who you are →