Advanced genetics & diagnostics
When standard testing isn't enough — the genetic markers and specialist tests worth considering if lipids stay stubbornly high despite full lifestyle and supplement intervention.
Deep-dive into genetic and advanced testing Genetic risk markers, polygenic scores, advanced imaging and specialist lipid analysis
Advanced genetic considerations
Two people with the exact same elevated LDL level can have drastically different lifetime risks for a cardiovascular event based entirely on their genetics. Genetic tests do not replace routine blood panels; they provide a completely different dimension of data. While a blood test tells you how fast atherosclerosis is likely progressing right now, a genetic test reveals your lifetime cumulative exposure and whether your foundational lipid-clearing machinery is broken.
The markers below are clinically significant when evaluating routinely high lipids — particularly if conventional intervention isn't producing the expected response. Many can be found in consumer genetic services (23andMe, Nebula Genomics, Living DNA) or through specialist lipid clinics in the NHS or private settings.
Monogenic "Big Three" — Familial Hypercholesterolaemia (FH)
Single-gene mutations responsible for FH — the most common inherited cause of severe hyperlipidaemia (~1 in 250 UK adults, though ~80% remain undiagnosed). Testing is clinically vital because carriers have likely been exposed to massively elevated cholesterol since birth, with severe risk of early-onset ischaemic heart disease. NHS FH genetic testing is available with strong family history (premature CHD or known FH relative).
The most common genetic culprit. A mutation here means your liver physically lacks the receptors necessary to pull LDL particles out of your bloodstream — leaving them trapped in circulation to oxidise and form plaques. Accounts for ~80-85% of identified FH cases. Heterozygous carriers typically have untreated LDL of 5-10 mmol/L from birth; homozygous (very rare) is much higher and life-limiting without intensive treatment.
The APOB protein on each LDL particle is the "key" that allows it to dock with hepatic LDL receptors. Specific APOB mutations (e.g. R3527Q/R3500Q) change the shape of the key — the LDL particle cannot dock correctly, rendering normal receptors effectively useless. Accounts for ~5-10% of FH cases; clinical pattern slightly milder than LDLR FH but still substantial.
Normal PCSK9 destroys LDL receptors as part of routine turnover. Gain-of-function mutations cause the body to hyperactively destroy its own LDL receptors, crippling hepatic cholesterol clearance. Rare (<5% of FH) but clinically severe. Conversely, loss-of-function PCSK9 variants are protective and led to the development of PCSK9 inhibitor drugs (alirocumab, evolocumab, inclisiran).
Polygenic risk — the "death by a thousand cuts"
Most elevated lipids aren't caused by a single catastrophic gene defect like FH. They're driven by the combined effect of thousands of small genetic variations scattered across the entire genome. Polygenic Risk Scores (PRS) aggregate these into a single risk profile — and the result can be surprising.
Aggregates ~80-200 small genetic variants that each nudge LDL slightly. Genetic background acts as an independent threat multiplier: individuals with a high PRS and merely "borderline" LDL-C can face a higher relative risk for coronary artery disease than individuals with "very high" LDL-C but a low genetic risk. Critical for explaining why some patients with modest lipid elevations still progress aggressively.
Like LDL, there's a polygenic risk architecture for triglycerides — explaining the genetic variance in dietary response. Why some individuals experience sharp, dangerous spikes in TG from a moderate dietary shift while others remain stable. Common variants involved include rs964184 (APOA5/ZPR1), rs13702 (LPL), rs505151, rs662799 (APOA5), and rs2266788. Your CGM patterns + TG-PRS together explain most of the variance in metabolic response to diet.
The single strongest common genetic variant for coronary artery disease — independent of the lipid pathway entirely. Acts via vascular wall biology (CDKN2A/CDKN2B antisense regulation). Each risk allele increases CAD risk ~20%; homozygous carriers (~25% of population) have ~50% increased lifetime risk regardless of cholesterol numbers. The classic "low cholesterol but CV event" patient often has 9p21 risk genotype.
Genetic determinants of Lp(a) levels — different from measuring Lp(a) itself. KIV-2 copy number variation determines particle size and circulating concentration. rs10455872-G and rs3798220-C are strongly associated with elevated Lp(a) and accelerated atherosclerosis. Genotyping reveals lifetime hazard from birth, not just current circulating level.
Pharmacogenomics — statin response
If statin intolerance has been an issue — or if you're considering starting one — pharmacogenomic testing can predict your response before a single dose. SLCO1B1 is one of the few PGx variants with clear, actionable clinical guidance (CPIC, NICE-acknowledged).
Encodes the OATP1B1 transporter that moves statins into liver cells. The *5 (C) variant impairs statin uptake by the liver, so more statin remains in muscle tissue — the direct cause of statin-induced myopathy. CC homozygotes have ~17× increased myopathy risk; TC heterozygotes ~4×. CPIC guidance recommends avoiding simvastatin entirely in C-allele carriers and dose-reducing other statins. The single most clinically actionable PGx test for cardiovascular medicine.
The hereditary haemochromatosis genes. C282Y homozygosity causes classical iron overload disease (~1 in 200 of UK population). H63D is milder but H63D/H63D and compound heterozygotes (C282Y/H63D) still cause subclinical iron accumulation. Elevated iron is independently atherogenic — it catalyses LDL oxidation and ferroptosis in arterial walls. Anyone with persistently elevated ferritin should be HFE-tested; iron-free multivitamins become essential.
The vitamin D receptor sits on pancreatic beta cells and influences insulin secretion — making receptor variants candidates for modifying who benefits from vitamin D supplementation. A 2026 secondary analysis of the D2d trial (4000 IU/day vs placebo in prediabetic adults) found that only the ~70% of participants carrying the ApaI CC or AC genotype showed a meaningful reduction in T2D progression (19%, HR 0.81), while ApaI AA carriers (~30%) showed no benefit at all (HR 1.02). The original D2d trial as a whole missed statistical significance — splitting by genotype revealed the responders.
Caveats: single secondary analysis, not yet replicated. BMI partially confounds the effect (D2d showed benefit only in BMI < 30). The VITAL trial PI has said they're "considering" genotyping their cohort to test the same hypothesis. Treat as research-grade, not actionable yet. Useful framing for anyone with persistently low vitamin D + prediabetes + not responding to standard supplementation — discuss with your GP whether higher-dose trial is appropriate.
Extreme hypertriglyceridaemia genetics
In cases of extreme, routinely unmanageable triglycerides (>10 mmol/L), mutations in fat-clearing enzymes can cause Familial Chylomicronaemia Syndrome (FCS) — a rare condition where the body fundamentally lacks the machinery to clear dietary fat.
Encodes the enzyme that breaks down triglycerides from chylomicrons and VLDL particles in the bloodstream. Loss-of-function LPL mutations cause FCS — triglycerides routinely 10-100+ mmol/L, recurrent pancreatitis, eruptive xanthomas. Onset typically in childhood. Common heterozygous LPL variants (less severe) influence dietary fat response in the general population.
APOC2 is the essential co-activator for LPL. APOA5 modulates LPL activity. GPIHBP1 anchors LPL to capillary walls. LMF1 helps LPL mature. Mutations in any of these can produce FCS-like phenotypes through different mechanisms. The complete FCS workup tests all five (LPL + APOC2 + APOA5 + GPIHBP1 + LMF1).
Advanced diagnostic tests
Beyond the standard panels, several specialist tests can either directly visualise atherosclerosis or measure biomarkers that capture vascular biology beyond cholesterol numbers. These are most useful when lipid numbers and risk calculators give conflicting signals, or when you want to know what's actually happening in your arteries rather than predicting from blood markers alone.
Direct imaging — what's actually in your arteries
A non-contrast CT scan that directly measures calcified plaque in the coronary arteries. Result is an Agatston score: 0 = essentially zero CV risk over the next 5-10 years regardless of cholesterol numbers; 1-100 = mild; 100-400 = moderate; >400 = significant disease. Arguably the single most clinically useful CV imaging test. Low radiation (~1 mSv). The strongest case for "delaying or avoiding statins" is a zero CAC score in an otherwise borderline patient. See Section 10 for the full action plan if your CAC is above zero, the calcium paradox, and the K2/EPA stabilisation strategy.
Ultrasound measurement of the inner wall thickness of the carotid arteries. Detects atherosclerosis earlier than CAC because thickening precedes calcification. No radiation, repeatable. Useful for younger patients (<45) where CAC is typically zero even with active disease. Normal <0.75mm; elevated >1.0mm.
Tests how well your arteries dilate in response to blood flow demand — endothelial dysfunction is the earliest stage of atherosclerosis, often present years before plaque forms. EndoPAT (Reactive Hyperaemia Index) uses fingertip sensors; Flow-Mediated Dilation (FMD) uses brachial artery ultrasound. Functional, reversible — improves rapidly with diet, exercise, and Mediterranean intervention.
Advanced lipid particle analysis
Nuclear magnetic resonance separates lipoproteins by particle size and number. Reports LDL particle count, small dense LDL (sdLDL), IDL, large HDL, and the LP-IR insulin resistance score. Particularly useful when LDL-C and apoB give conflicting signals, or when small-dense LDL pattern is suspected (more atherogenic than equivalent large-buoyant LDL). Resolves the "metabolically healthy" vs "metabolically dyslipidaemic" question.
A single 0–100 number derived from the NMR profile above, weighting six lipoprotein parameters — large VLDL particles and VLDL size, small LDL particles and LDL size, large HDL particles and HDL size. Higher is more insulin resistant; a score of 68 or above is the cut-off this card means — the one derived from the PREVEND cohort and quoted on the laboratory's own test information. Be aware that a second figure circulates: some reports flag at 45. If your report is flagged at 45 and you are reading a source that says 68, you are not looking at the same threshold, and it is worth asking which the flag on your own report was set to before drawing any conclusion from it. It is a genuine, validated test, and what it is validated for is the part that gets lost.
Its strongest evidence is for predicting diabetes, not heart attacks. In the Women's Health Study (25,925 women, median 20 years) the hazard ratio for incident type 2 diabetes was 1.95 per standard deviation; in PREVEND the top quartile against the bottom gave a hazard ratio of 10.18, falling to 3.02 once clinical risk factors were accounted for. Those are the large numbers that circulate — and they are diabetes numbers.
For cardiovascular disease the manufacturer's own position is more modest: associations with subclinical atherosclerosis and incident cardiovascular disease comparable to HOMA-IR. Comparable to a decades-old calculation from fasting glucose and insulin — not superior to apoB, which is the comparison that matters and which nobody makes. It correlates with HOMA-IR only moderately (r = 0.51), so it is not simply a cheaper HOMA-IR either.
And much of what it tells you is already on your panel. In the same Women's Health Study analysis, adjusting for HbA1c, CRP, triglycerides, HDL-C and LDL-C attenuated the hazard ratio from 1.95 to 1.41. It is built from lipoproteins, so a large share of its signal is the triglyceride and HDL information you already have — which is what the TG/HDL ratio is a crude version of. Graded Moderate for insulin resistance and diabetes risk, and Emerging for adding anything to cardiovascular risk assessment beyond apoB, HbA1c and a standard panel.
Directly measures the damaged, oxidised form of LDL — the actual species that gets phagocytosed by macrophages and becomes foam cells in plaque. ApoB tells you how many "bullets" you have; oxLDL tells you how many have already been "fired" into your arterial wall. Elevated oxLDL predicts cardiovascular events independently of total cholesterol or LDL-C.
Vascular-specific inflammation
Lipoprotein-associated phospholipase A2 — an enzyme produced in arterial wall inflammation specifically, unlike hs-CRP which is a non-specific systemic inflammation marker. Elevated Lp-PLA2 means active vascular wall inflammation and increases stroke risk independently of LDL. The "PLAC test" is the standardised commercial form.
An enzyme released by white blood cells embedded in vulnerable, unstable plaque. Elevated MPO indicates plaque inflammation and predicts plaque rupture — distinct from how much plaque you have. Useful for stratifying patients with established CV disease into "stable" vs "unstable plaque" categories.
A gut-microbiome-derived metabolite produced when gut bacteria metabolise carnitine (from red meat) and choline (from eggs, fish). Elevated TMAO is associated with accelerated atherosclerosis and increased CV event risk, independent of cholesterol. Modifiable via dietary shift (more plant-based, less red meat) and gut microbiome interventions.
Sphingolipid species (Cer16:0, Cer18:0, Cer24:1) that accumulate in metabolic dysfunction. The Mayo Clinic Ceramide Coronary Score (MI score) is a validated 4-marker panel that independently predicts major adverse cardiac events better than LDL in several cohorts. Reduced by Urolithin A (Mitopure) per 2025-2026 trials — see Tier 3 longevity supplements.
Insulin & metabolic precision
HbA1c is a 3-month average and lags real metabolic change by 6-12 weeks. Fasting insulin captures insulin resistance years before HbA1c rises. HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) combines fasting glucose and fasting insulin into a single metric: optimal <1.0, suboptimal 1-2, insulin resistant >2. Critical for catching metabolic dysfunction in its earliest, most reversible stage.
Measures how much insulin the pancreas is producing (unlike serum insulin which includes administered insulin). Elevated C-peptide with normal glucose indicates compensatory hyperinsulinaemia — the pancreas working hard to maintain normoglycaemia in the face of resistance. Often the earliest detectable metabolic abnormality, preceding both HbA1c rise and fasting glucose rise by years.
The verdict: blood tests vs genetics vs imaging
Standard lipid blood tests are necessary to track the real-time success of your dietary, supplementary, or pharmacological interventions. However, if your lipids remain stubbornly high despite strict lifestyle control, genetic testing is not just meaningful — it is the definitive tool to uncover whether you are fighting a biological inevitability that requires specialised medical intervention (like a PCSK9 inhibitor, inclisiran, or — in extreme cases — apheresis).
And in the middle, between blood biomarkers and DNA, lies direct imaging. A CAC score of zero in a "high-risk" 50-year-old with elevated LDL is a far stronger argument for delaying or avoiding lipid medication than any blood test could provide. Conversely, a high CAC score in a "low-risk" patient is a non-negotiable call to action. Together, the three layers — biomarkers, genetics, and imaging — give you the complete picture: how fast the fire is burning, how flammable the building is, and whether anything is already on fire.
A note on Lp(a): what to do while there is no drug for it
Lp(a) is the one result on this page that comes with no treatment. Statins do not lower it and can nudge it slightly up; ezetimibe does nothing; PCSK9 inhibitors lower it by perhaps 20–25% as a side effect of doing something else. So the reasonable question after a high result is what it is for. Watch this space, and here is the date to watch. Lp(a)HORIZON, testing whether lowering Lp(a) with pelacarsen actually reduces cardiovascular events, is the first outcome trial this field has ever had; its readout is expected before the end of 2026. Until it reports, everything about treating Lp(a) is a hypothesis about a risk marker — and this table will change when it does. Checked 15 August 2026.
Three things, none of them a prescription for the Lp(a) itself.
- It reprices everything else. A high Lp(a) is risk you cannot remove, which raises the value of the risk you can. It strengthens the case for treating apoB harder and earlier, and for taking blood pressure, smoking and metabolic health more seriously than the headline risk score alone would suggest.
- It is a reason to look, not just to calculate. A high Lp(a) with an unremarkable risk score is one of the better arguments for imaging — a CAC score answers whether the risk has already turned into anything.
- It is inherited, so it is a family finding. Because it is set by a single gene and essentially fixed for life, first-degree relatives have roughly a one-in-two chance of carrying the same elevation. Cascade testing is the highest-yield thing a high Lp(a) result produces, and it is the step most often skipped.
The drugs that are coming, and where they actually are
Four agents are in phase 3 outcome trials, all lowering Lp(a) dramatically in earlier work — roughly 80% for pelacarsen, over 90% for the RNA-silencing agents. None is approved, and none has yet shown that lowering Lp(a) reduces cardiovascular events. That is the entire question, and it is unanswered.
| Agent | Type and dosing | Outcome trial and status |
|---|---|---|
| Pelacarsen | Antisense oligonucleotide, monthly injection | Lp(a)HORIZON — the furthest advanced. Guidance pointed at a mid-2026 readout; as of writing that window has passed with no result published. |
| Olpasiran | siRNA, injection every 12 weeks | OCEAN(a)-Outcomes, around 7,300 patients with established disease and Lp(a) at or above 200 nmol/L. Completion expected end of 2026. |
| Lepodisiran | siRNA, potentially twice yearly | ACCLAIM-Lp(a), a larger programme including higher-risk primary prevention. Still running. |
| Muvalaplin | Oral, once daily | MOVE-Lp(a). Earliest of the four. |
Why this page will not tell you to wait for them. Lp(a) is the strongest single genetic risk factor in cardiovascular medicine and it is a textbook case for treatment — and so was inflammation, until CLEAR SYNERGY and ZEUS. A biomarker being causal, heritable and modifiable does not guarantee that modifying it helps, and the first outcome trial is the thing that finds out. Emerging Until one reports, the three actions above are the whole of what a high Lp(a) should change.
Currency warning — read the date before you rely on this. This is the fastest-moving section on the site and a single readout will date the whole table. Lp(a)HORIZON's guided window closed on 30 June 2026 with nothing disclosed, and the ESC congress is weeks away. Reviewed August 2026; check the trial registries before quoting any of it.
The inherited conditions where the story is underdiagnosis
Everything else on this page is about gradual risk — how fast plaque accumulates over decades. This section is about a different shape of problem: uncommon inherited conditions that are largely silent, that concentrate in families, and where the first symptom is sometimes the last one. They belong on a prevention site for exactly that reason.
None of this is a reason for a well person with no family history to go looking. The trigger is almost always the same, and it is worth stating plainly.
The family history that should be acted on
- A sudden, unexplained death in a relative under about 40 — including deaths recorded as drowning, a single-vehicle car accident, or sudden infant death, all of which can be arrhythmic deaths that were never investigated as such.
- A relative with an inherited heart condition already diagnosed, in which case testing relatives is the standard of care rather than an unusual request.
- Fainting during exercise, or during a sudden fright or loud noise. Fainting after exercise stops is usually benign; fainting during exertion is not, and warrants assessment.
- Unexplained heart failure, pacemakers or transplants in relatives under 50.
Familial hypercholesterolaemia — the common one
Covered throughout this guide, but it belongs in this list because it is the clearest example of the pattern. It affects roughly 1 in 250 people, it is straightforward to detect from a lipid panel and a family history, treatment is cheap and highly effective, and the large majority of people who have it in the UK do not know. Where it is diagnosed, testing first-degree relatives is the highest-yield step in the whole of preventive cardiology — each has a one-in-two chance. Strong
Hypertrophic cardiomyopathy
Thickening of the heart muscle, usually from a mutation in a gene coding for part of the contractile apparatus, affecting somewhere around 1 in 500. Most people with it live a normal lifespan, and many are never diagnosed. It matters here because it is among the commonest causes of sudden cardiac death in young people and athletes, because it is detectable — an ECG is usually abnormal and an echocardiogram is definitive — and because risk can be estimated and, where high, reduced with an implantable defibrillator. Strong
The inherited arrhythmia syndromes
Long QT syndrome, Brugada syndrome and catecholaminergic polymorphic ventricular tachycardia are conditions of the heart's electrical system rather than its structure. The heart looks normal on a scan; the abnormality is in how it repolarises. Long QT is the commonest, at perhaps 1 in 2,000, and has a practical consequence beyond the diagnosis itself: a long list of ordinary medicines prolong the QT interval further, including some antibiotics, antihistamines, antipsychotics and antiemetics. Knowing you have it changes prescribing for the rest of your life. Strong
What assessment actually involves
Less than people expect. A resting ECG and an echocardiogram will identify most structural and many electrical conditions, and both are quick and non-invasive. Genetic testing follows a finding rather than replacing it, and is most informative when a specific variant has already been identified in an affected relative — at which point testing the rest of the family becomes a clear yes-or-no rather than a probability.
Two limits. Population screening of everyone is not recommended anywhere, and for good reason: these conditions are uncommon, ECG interpretation in young and athletic hearts produces a considerable number of false positives, and the consequences of a wrong label — sport withdrawn, insurance affected, a lifetime of anxiety — are not trivial. The case for testing rests on family history, not on curiosity. And a genetic test that finds nothing does not exclude the condition; a proportion of clearly affected families still have no identifiable variant.
Cascade testing from a known case is the efficient middle path, and it is what the UK does instead of population screening: once one person in a family is diagnosed, testing relatives has a high yield, a clear question and none of the false-positive problem that screening a whole population creates. It is also the step most often left undone.
In the UK these are managed through inherited cardiac conditions services, referral to which comes through a GP or cardiologist. If the trigger above applies to your family, that referral is the whole of what this section is asking you to consider.