The statin threshold just dropped 🧬

LONGEVITY LATEST ISSUE 24 · 19 AUGUST 2026

LONGEVITY LATEST

The Evidence-Based Edge on Living Longer and Better

Issue 24 · The Treatment End · 19 August 2026

WELCOME

👋 Welcome

Five per cent. That is the ten-year risk that now earns you a firm statin recommendation, down from seven and a half.

Last week was about measurement. This week is the other half. In March the new guideline moved the treatment line; in February the Lancet published the largest analysis ever assembled of what statins do to the people taking them. Both went almost unreported. I read both, and the short version is that the argument has moved on without most of us.

In this issue:

• 🔬 Top 3: the pill at roughly a pound a month, the £4 add-on almost nobody is offered, and the £45 one for people who genuinely can't take the first

• ⭐ Spotlight: 66 side effects on the label, 38 million adverse-event records, and the four that survived

• 🚨 Hype Check: the natural alternative that is a statin wearing a different label — and the vote in May that says so

• 📖 Deep Dive: five drugs, one lever, and why your arteries can't tell them apart

THIS WEEK'S ANALYSIS

🔬 Top 3 Interventions Under the Microscope

Three drugs rather than three tests this week. Two earn an A and one earns a B+ — and, awkwardly, the B+ is the one costing forty-five times what the A does. Which makes the ranking a question of sequence rather than enthusiasm.

1. Statins — Evidence Grade: A

What it is. A statin blocks HMG-CoA reductase, the rate-limiting enzyme in the liver's own cholesterol line. Short of supply, the liver builds more LDL receptors and starts pulling apoB particles out of your blood to make up the difference. The drug doesn't scrub your arteries. It makes your liver hungrier for the lorries from last week's issue.

Human evidence. The Cholesterol Treatment Trialists pooled individual data from 27 trials and roughly 175,000 people. Among those at under 10% five-year risk — the group guidelines had been leaving alone — each 1 mmol/L removed prevented about 11 major vascular events per 1,000 over five years.

That was 2012. In February a meta-analysis in the Journal of Clinical Lipidology took 13 outcomes trials of wholly or mostly primary-prevention patients — close to 98,500 people — and found a 30% relative risk reduction per mmol/L. Thirty, against roughly 20% in people who already have disease. The benefit grows proportionally as baseline risk falls, exactly as last week's cumulative-exposure argument predicts.

Cautions. Two effects are real, and neither is the famous one. New-onset diabetes, concentrated almost entirely among people already close to the diagnostic line; and a small excess of muscle symptoms, which gets its own section below. Simvastatin and atorvastatin run through CYP3A4, so clarithromycin, itraconazole and several antivirals matter. Rosuvastatin and pravastatin largely dodge it.

Takeaway. Grade A, and the case is strongest exactly where doctors have been most reluctant. The argument was never whether statins work. It was who gets offered one, and that answer just widened by several million people.

2. Ezetimibe — Evidence Grade: A

What it is. Different door, same room. Ezetimibe blocks NPC1L1, the transporter pulling cholesterol out of the gut — both what you ate and the larger amount your own bile delivers. Less reaches the liver, which responds exactly as it does to a statin. Statins turn down production. Ezetimibe turns down imports.

Human evidence. IMPROVE-IT randomised 18,144 people after an acute coronary syndrome to simvastatin with or without ezetimibe. At seven years, 32.7% against 34.7% — two percentage points, hazard ratio 0.936. Modest, and it took a decade to deliver. RACING is the one that changed my mind about sequencing: 3,780 patients with established disease, randomised to a moderate statin plus ezetimibe or a high-intensity statin on its own, and at three years 9.1% against 9.9% — non-inferior, with fewer people quitting over side effects. Two moderate levers matched one aggressive one, and were easier to keep pulling.

Cautions. Alone it moves LDL cholesterol by less than a fifth. An addition, not a substitution, unless statins are genuinely off the table.

Takeaway. Grade A, £4.64 for a month on the current NHS tariff, and one of the most under-offered drugs in the cabinet. If you are on a statin and still above target, this is the conversation to have.

3. Bempedoic acid — Evidence Grade: B+

Fair warning: this one costs forty-five times what the first does, and that ratio is the whole argument about where it belongs.

What it is. An ATP-citrate lyase inhibitor, two steps upstream of a statin on the same chain and ending at the same receptors. The clever part is delivery: it arrives as a prodrug needing an activating enzyme the liver has and skeletal muscle doesn't, so it was engineered around the exact complaint that stops people taking statins.

Human evidence. CLEAR Outcomes randomised 13,970 people who couldn't or wouldn't take a statin. LDL cholesterol fell 21%, major cardiovascular events 13% (HR 0.87). In the primary-prevention third the hazard ratio was 0.70 — 43 treated to prevent one event. I'd want a dedicated trial there first: a subgroup that outperforms its parent trial usually means a wide interval, and this one runs from 0.55 to 0.89.

Cautions. Uric acid rises and gout follows in some people. Gallstones and tendon rupture appear in the safety data. And £45.36 a month against roughly a pound for atorvastatin.

Takeaway. Grade B+, and the reasoning matters: one outcomes trial, large and clearly positive — but one, where the two drugs above it have several apiece pointing the same way. That gap, plus the price, is why this sits third. The drug for people who have genuinely failed the other two — a much smaller group, as the next section explains, than currently believes it belongs there.

SPOTLIGHT

⭐ Spotlight: The Side-Effect List That Lost 62 Items

I went in expecting a modest tidy-up of the statin safety literature. What arrived in February was closer to a demolition.

The Cholesterol Treatment Trialists took 19 double-blind placebo-controlled trials — 123,940 participants, median follow-up four and a half years — built from more than 800 datasets and 38 million adverse-event records. Then they went through the 66 conditions listed on statin labels as possible undesirable effects and asked, one at a time, whether the randomised data showed a causal excess.

Sixty-two of the sixty-six showed no causal relationship with the drug at all.

Pros. Four survived: liver-enzyme rises, abnormal liver function tests, oedema, and changes in urinary composition — alongside muscle symptoms and new-onset diabetes, both already established elsewhere. What did not survive is the list people actually quit over. Cognitive impairment, depression, sleep disturbance, peripheral neuropathy: no causal signal for any of them.

The muscle question has its own history. The CTT's 2022 analysis of almost 155,000 people found 14 of every 15 muscle-symptom reports on a statin weren't caused by the statin, with the excess gone after year one. SAMSON then put 60 people who had abandoned statins through twelve blinded months of atorvastatin, placebo and empty bottles. Symptom intensity: 16.3 on the statin, 15.4 on placebo, 8.0 on nothing. Half restarted afterwards.

Cons. Three caveats I won't bury. Trial populations skew towards people who already tolerated the drug, so this reads better as a case for continuing than for starting. Adverse-event recording in 1990s trials was never designed for this question. And “no causal excess across 123,940 people” is a statement about populations, not individuals. The fifteenth person is real, and telling them it's imaginary is both wrong and useless.

HYPE CHECK

🚨 Hype Check: Red Yeast Rice, the “Natural” Statin

The Hype: Sold across British health shops as the natural alternative for people who'd rather not take a statin — Simply Supplements at 2.5 mg of monacolin K a capsule, HealthAid, Health4All at 600 mg, generally £10 to £20 a month. Some labels state outright that monacolin K is also known as lovastatin: admirable candour, or a very quiet confession.

The Evidence: Monacolin K in its lactone form is lovastatin. Not similar to it — identical. EFSA said so in 2018, which is why the EU capped supplements below 3 mg a day in 2022 and why member states voted on 13 May 2026 to ban monacolins outright. And when the Cleveland Clinic's SPORT trial ran a 1,200 mg red yeast rice product against rosuvastatin 5 mg in 190 adults, the statin cut LDL cholesterol 37.9% and the red yeast rice did not differ from placebo.

Why It's Misleading: Start with the label, because I went looking for the British version of the 3 mg cap and there isn't one. It binds in Northern Ireland but was never adopted in Great Britain, so one retailer's own packaging tells Northern Irish customers to take one capsule a day and British customers up to four.

Same brand, same bottle, four times the dose depending on which side of the Irish Sea you open it.

Then the contents. Across 37 European products, citrinin — a mould by-product that damages kidneys in animals — turned up in every one, at up to 25,100 µg/kg against a ceiling of 100. Declared monacolin K was wrong both ways, from 83% under to 266% over.

SUPERFOOD

🥣 Superfood Spotlight: Pulses, and a Number That Cuts Both Ways

Beans, chickpeas, lentils, peas. The pooled evidence runs to 26 randomised trials and 1,037 people at a median 130 g a day — one tin, near enough. LDL cholesterol fell 0.17 mmol/L. Here's the part I'd rather tell you than skip: apoB and non-HDL, the two numbers last week's issue told you to trust, didn't shift significantly, though far fewer trials measured them. Pulses do something real to the marker we spent an issue demoting, and we don't yet know what they do to the ones we promoted. A tin of chickpeas is 55p. At that price the evidence needn't be spectacular. It needs to be real, and it is.

DEEP DIVE

📖 Deep Dive: Five Drugs, One Lever

Three drugs, three unrelated targets: an enzyme in the liver cell, a transporter in the gut wall, a rung two steps up the synthesis chain. Your arteries cannot tell them apart.

I'd assumed that was a convenient simplification until I read the genetics. Ference's factorial study naturally randomised 108,376 people by the variants they were born with — some in NPC1L1, ezetimibe's target, some in HMGCR, the statin's. The NPC1L1 group: 2.4 mg/dL lower LDL cholesterol, 4.8% less coronary heart disease. The HMGCR group: 2.9 mg/dL, 5.3% less. Per unit removed, indistinguishable.

So the mechanism isn't the point. The artery only counts what came off the number, multiplied by the years it stayed off.

👉 Read the Deep Dive: Five Drugs, One Lever →

BIOHACKING CORNER

🌡 Biohacking Corner: Do One Thing This Week

Work out your ten-year cardiovascular risk before you do anything else. In the UK, QRISK3 is free, takes about four minutes, and needs only your last blood pressure and cholesterol readings. Everything below hangs off that one number.

1. Between 3 and 5%? The new conversation band. Not an automatic prescription — a discussion, with family history and Lp(a) on the table.

2. Above 5%? You now meet the threshold for a firm recommendation. If nobody has raised it with you, raise it with them.

3. On a statin and still above target? Ask about ezetimibe by name. Four pounds sixty-four a month, and RACING says a moderate statin plus ezetimibe matches a high-intensity statin alone.

4. Stopped one over side effects? Half of SAMSON's participants restarted once they'd seen their own blinded data. Ask about a structured re-challenge — different statin, lower dose, alternate days — rather than trying again alone and hoping.

5. Retest at eight to twelve weeks, not three. Lipids need time. A three-week reading tells you about last weekend, not about the drug.

I ran QRISK3 on myself before writing this, expecting to waste four minutes. It came back higher than I'd assumed, for a reason I could have predicted: age is the heaviest variable in the model, and I keep forgetting to age.

READER PULSE

📊 Reader Pulse

The fitness poll, back as promised. Forty-six per cent of you train mainly for strength, 31% for endurance, 18% chose “whatever my knees still permit”, and 5% don't train and would like us to stop asking. The strength number tracks the Issue 21 evidence almost exactly — either good reading or good self-selection.

This week: do you know your ten-year cardiovascular risk score?

• “Yes, and I've checked it within the last year.”

• “I knew it once. Years ago.”

• “No — and I've just worked it out.”

• “No, and I would rather not.”

Vote at longevitylatest.com/poll-24.

CLOSING

🎯 Closing

This is where the plumbing stops being about knowing and starts being about doing. The measurement was free. The treatment, if you need it, is about a pound a month. The expensive part was always the twenty years of not asking — and I spent most of mine not asking.

Next week, the pressure inside the pipes: whether home readings beat the clinic's, what the evidence says about treating to 120, and why the commonest blood pressure error happens in the thirty seconds before the cuff inflates.

Go and find your number.

Stay curious and stay healthy!

— Christian Thomsen, Editor

Longevity Latest is published weekly by FrontWave Media Ltd. The content is for educational purposes and does not constitute medical advice. Decisions about lipid-lowering treatment depend on your full risk profile and should be made with a clinician. Do not start, stop or alter any prescribed medication on the basis of this newsletter. If you have known cardiovascular disease, familial hypercholesterolaemia, diabetes, liver or kidney disease, or you are pregnant or breastfeeding, speak to your doctor before making any change. Consult your physician before starting any new regimen.

© 2026 FrontWave Media Ltd · Longevity Latest 1