Your lipid panel has five or six numbers on it and they are not equally important. Which one carries the risk, which one is a message about something else, and how far diet and training actually move them.
Kratos Natural · Domain: Blood work & testing
1. Why does cholesterol matter?
Start with the plumbing, because the rest of the guide follows from it.
Fat does not dissolve in water. Blood is mostly water. So fat travels wrapped: a droplet of fat and cholesterol inside, a shell of protein and phospholipid outside, and the shell is what lets the whole thing move through plasma. Those wrapped droplets are lipoproteins, and their names are on your report. LDL. HDL. VLDL.
Size and cargo differ between them. The line that decides who gets atherosclerosis is elsewhere. Some of these particles carry a protein called apolipoprotein B, and some do not, and the ones that do are the ones that get into an artery wall and stay there. LDL carries it. VLDL carries it. Remnants and Lp(a) carry it. HDL does not.
Now the useful part. A particle carrying apoB carries exactly one of them. One shell, one apoB, every time. So a measurement of apoB is a headcount of the particles that can cause the disease.
Your report does not give you that headcount. It gives you the weight of cholesterol those particles happen to be carrying between them, because measuring cholesterol was the practical option when these panels were designed and the two usually track each other closely enough. Usually. Somebody whose particles are each carrying slightly less cholesterol than average has a respectable LDL and more particles than that number implies, and that person is the reason this chapter exists.
2. What is ApoB?
The cleanest way to settle an argument between correlated measurements is to put them in the same model and see which one survives. That has been done, at scale.
Researchers took 389,529 people from the UK Biobank who were on no lipid-lowering treatment, and 40,430 people with established atherosclerosis taking a statin in the FOURIER and IMPROVE-IT trials, and tested apoB, non-HDL cholesterol, LDL cholesterol and triglycerides against later heart attacks.
One at a time, all of them predicted. Assessed together, only apoB remained, at a hazard ratio of 1.27 per standard deviation (95% CI 1.15 to 1.40). Same answer in the people who already had the disease as in the people who did not[1].
There is a second result in that paper worth more than it first appears. Once apoB was accounted for, the ratio of triglycerides to LDL cholesterol carried no further information about risk. That ratio is a proxy for what kind of particles somebody has, LDL versus the triglyceride-rich ones. So for a given headcount, the mix stopped mattering.
Count, not cargo, and not type.
The editorial the journal ran alongside it is titled The Debate Is Over[2]. That is an opinion piece rather than evidence, and it is quoted here because it is a fair description of where lipidology has landed.
So why is it not on your report? Because nobody ticked the box. ApoB is a standard assay, every lab runs it, it costs a few euros, and it needs no fasting. Ask for it once alongside the usual panel and you will know whether your LDL is telling you the truth about your particle count.
Which is why apoB sits at the top of the lipid panel here as a core value, with under 1.0 g/L from the ESC/EAS guideline. That threshold is written for average risk. Higher risk moves it down, and where your line sits is a conversation with a doctor rather than a number in a guide.
3. What is LDL cholesterol?
The strongest claim in this guide, and the one with the most behind it.
The European Atherosclerosis Society did not run a new study. It assembled the existing ones: more than 200 prospective cohorts, plus Mendelian randomisation studies, plus randomised trials, together over 2 million people, over 20 million person-years and over 150,000 cardiovascular events.
The point of assembling those three in particular is that they fail differently. Cohort studies are vulnerable to confounding. Genetic studies are not, because alleles are dealt at conception, but they cannot tell you what happens when you intervene in an adult. Randomised trials can, but they run for five years rather than fifty. Each covers the others' weakness, and all three produced the same dose-dependent, log-linear relationship with no threshold anywhere in it. The conclusion runs to one sentence: LDL causes atherosclerotic cardiovascular disease[3].
What that buys, from the trial side. The Cholesterol Treatment Trialists pooled individual participant data from 26 trials and 170,000 people. Per 1 mmol/L of LDL reduction, major vascular events fell by roughly 22% (rate ratio 0.78, 95% CI 0.76 to 0.80). All-cause mortality fell 10% per mmol/L. No threshold appeared within the range studied, including among people already under 2 mmol/L on the comparison regimen[4].
4. Why does exposure over years matter most?
This chapter is separate from the last one because the idea in it is the one most often missed, and it is the reason a healthy thirty-year-old should read a lipid panel at all.
Ference and colleagues studied people carrying variants in any of six genes that left them with slightly lower LDL for their entire lives. No drug, no diet, no decision: an allele. Per mmol/L of lifelong lower LDL, coronary heart disease risk was 54.5% lower (95% CI 48.8 to 59.5) across 312,321 participants. That is roughly three times the benefit the same one mmol/L delivers when it starts in someone's sixties[5].
The quantity that damages an artery is not a concentration. It is a concentration multiplied by the years it has been there. A modest number held for four decades outperforms an aggressive intervention begun late, and neither your report nor any risk calculator shows you that product.
The target shown on this dashboard is under 3.0 mmol/L, from the ESC/EAS guideline, with the same caveat as apoB: more risk elsewhere, lower line.
5. What are triglycerides?
Triglycerides are fat in transit rather than fat parked. Your liver and your gut package them into large particles and send them out to be unloaded, and what is left over once the cargo has gone is called a remnant: small, apoB-bearing, and now the right size to lodge in an artery wall.
The reference review here sets two boundaries. Between 2 and 10 mmol/L cardiovascular risk is increased. Above 10 mmol/L the risk of acute pancreatitis is added to it. The same review is straightforward that randomised evidence for lowering triglycerides specifically was, at the time, thin[6].
That evidence has since been assembled, and the answer is qualified rather than clean. A meta-regression covering 374,358 people and 46,180 major vascular events, drawing on fibrate, niacin, omega-3 and statin trials, found a rate ratio of 0.84 (95% CI 0.75 to 0.94) per mmol/L of triglyceride reduction, independent of LDL. Then REDUCE-IT turned out to be a significant outlier and a strongly influential trial. Removed, the estimate weakened to 0.91 and lost significance[7].
So a raised triglyceride number is genuine risk, worth less per mmol/L than LDL, and its greater value is as information. It goes up with visceral fat, with alcohol, with refined carbohydrate, and with insulin resistance that has not yet shown up in your glucose. Of everything on the panel it is also the number that responds fastest, which makes it the one worth re-measuring after a change.
The threshold shown here is under 1.7 mmol/L, which is a fasting figure. Chapter 9 explains why that matters.
6. What is HDL cholesterol?
Every observational dataset agrees that people with high HDL have fewer heart attacks. That is not in dispute. What was in dispute is whether the HDL is doing any of the work, and the way that got answered deserves describing, because the same logic appears three times in this guide.
Alleles are assigned at conception, before anything about your life has happened to confound them. So a variant that raises HDL for life is, in effect, a randomised trial that ran for seventy years and cost nothing.
In 20,913 heart attack cases and 95,407 controls, carriers of the LIPG 396Ser variant had HDL 0.14 mmol/L higher than non-carriers and were unremarkable on every other lipid and non-lipid risk factor. On the observational relationship, that much extra HDL should have shown up as roughly 13% fewer heart attacks. It showed up as none: odds ratio 0.99 (95% CI 0.88 to 1.11). A score built from fourteen HDL-specific variants gave the same null. Run as a positive control on LDL, the identical method produced a large and significant effect, which rules out the method as the explanation[8].
Read HDL as an instrument, then, and not as a lever. A low HDL is genuine information, usually about the same metabolic picture that raises your triglycerides. Raising the number itself has been the graveyard of several drug programmes.
So the app prints thresholds for it, above 1.0 mmol/L for men and 1.2 for women from ESC/EAS, and offers no advice anywhere on how to push it up.
7. What is Lp(a)?
An LDL particle with an extra protein, apolipoprotein(a), bonded to it. How much of it you make is set almost entirely by the gene you inherited. Diet barely touches it. Training does not touch it. It is close to a fixed property of you.
The 2022 European consensus statement concludes that epidemiological and genetic evidence in hundreds of thousands of people supports a causal and continuous relationship with cardiovascular outcomes across ethnicities, and that raised Lp(a) remains a risk factor even at very low LDL cholesterol. It is also associated with calcification of the aortic valve. The panel recommends measuring it at least once in adulthood[9].
Once, because the answer does not change. That is why this app schedules it at a ten-year interval instead of six months. It is not a value you follow. It is a fact about you that you either know or do not.
The statement is equally clear about the limits. No therapy that specifically lowers Lp(a) has yet been shown to prevent events, and those trials are ongoing. Until they report, the recommendation for someone with a high result is to manage every other risk factor earlier and harder than their overall risk score alone would suggest. A high Lp(a) does not change your options. It changes the urgency of the ones you already had.
The threshold used here is 75 nmol/L. Check your units before comparing anything: labs report Lp(a) in nmol/L or in mg/dL, and converting between them is not a fixed multiplication, because the particle's mass varies with the size of the apo(a) protein.
8. What about total cholesterol and ratios?
Total cholesterol adds the cholesterol inside your apoB particles to the cholesterol inside your HDL. It sums a hazard and a bystander, which is why it is the weakest line on the report: two people with an identical total can have very different particle counts. The ESC/EAS figure of 5.0 mmol/L is shown here because it appears on every report somebody will bring to this app, not because it is worth steering by.
Non-HDL cholesterol is total minus HDL, which leaves everything carrying apoB. It costs nothing, needs no extra assay, and you can work it out from a report you already have. It is clearly better than LDL alone. It only lost when placed directly against apoB[1].
Ratios compress two facts into one and lose the ability to act on either. A total-to-HDL ratio that looks healthy because HDL is high, in someone whose apoB is also high, reads as reassurance and functions as a blind spot. Chapter 6 explains why the numerator and denominator do not deserve equal billing.
If you can only add one thing to a request form: apoB. Failing that, non-HDL. Failing that, LDL.
9. Do you have to fast before the test?
For decades, yes, twelve hours of it. For most people that requirement is gone, and the joint statement from the European Atherosclerosis Society and the European Federation of Clinical Chemistry says exactly how much is at stake.
Maximal mean changes one to six hours after habitual meals: triglycerides +0.3 mmol/L, total cholesterol −0.2, LDL −0.2, non-HDL −0.2. HDL, apoA1, apoB and Lp(a) do not move with fasting status at all. Non-fasting and fasting values vary similarly over time and predict cardiovascular disease comparably. The recommendation is routine non-fasting sampling, with a fasting sample considered when non-fasting triglycerides exceed 5 mmol/L[10].
Two consequences you will meet on your own report.
The flagging thresholds move with the sample. The same statement asks labs to flag non-fasting triglycerides at 2 mmol/L and fasting ones at 1.7. Judging a non-fasting result against a fasting threshold manufactures a problem.
Consistency beats fasting. If you want two results to be comparable, have them drawn the same way, at roughly the same time of day, ideally at the same lab. That is also why every value in this app carries the date it was taken. The trend is the useful object and it is only trustworthy if the points were placed the same way.
10. How much do diet and training change it?
Effect sizes are given here on purpose. Advice about cholesterol is usually delivered without them, and the honest summary is that lifestyle moves these numbers genuinely and by less than most people are led to expect.
Saturated fat. The Cochrane review pooled 15 randomised trials with 56,675 participants, every one running at least two years. Reducing saturated fat cut combined cardiovascular events by 17% (risk ratio 0.83, 95% CI 0.70 to 0.98). In primary prevention 56 people had to sustain the change for around four years to prevent one event. There was little or no effect on all-cause mortality. Meta-regression found that the size of the reduction in saturated fat predicted the size of the benefit, which accounted for most of the disagreement between trials. Replacing those calories with polyunsaturated fat or with carbohydrate both worked; for monounsaturated fat the data were too sparse[11].
Soluble fibre. Fifty-eight trials, 3,974 participants, at a median 3.5 g of oat beta-glucan per day: LDL down 0.19 mmol/L (95% CI 0.14 to 0.23), non-HDL down 0.20 mmol/L, and apoB down 0.03 g/L[12].
Small, real, and unusually well measured for a nutrition question, since apoB was an outcome rather than an afterthought. The dose is a substantial bowl of oats, around 60 to 80 grams, daily rather than often.
Aerobic training. Forty-eight datasets, 2,990 sedentary adults with at least three metabolic syndrome factors, minimum twelve weeks: total cholesterol 0.19 to 0.29 mmol/L lower, triglycerides 0.17 to 0.18 lower, LDL 0.12 to 0.20 lower, HDL 0.05 to 0.10 higher[13].
Hold that against chapter 3, which counts in whole mmol/L. Training shifts a lipid panel by around a fifth of one. Train for your blood pressure, your insulin sensitivity, your VO2max and your legs at eighty. The lipid effect is a small bonus and not the case for it.
Alcohol and visceral fat, for triglycerides specifically. No trial figure is quoted here, which is deliberate rather than careless: it is the mechanism from chapter 5, and in practice it is where most raised triglyceride results come from. Expect that number to respond within weeks. Do not expect the other three to.
11. Which results need a doctor?
Book an appointment rather than a diet:
- LDL above 5 mmol/L. The threshold at which the consensus statement in chapter 9 asks for heterozygous familial hypercholesterolaemia to be considered[10]. An inherited receptor fault, present from birth, and considerably more common than most people assume.
- Triglycerides above 10 mmol/L. The only value in this guide that is dangerous in the short term. The risk is acute pancreatitis[6], and it is not something to raise at a routine review in a month.
- A heart attack or stroke in a parent or sibling before 55 in men or 60 in women.
- A high Lp(a), particularly alongside any of the above.
- Existing cardiovascular disease, or diabetes. Different and lower targets apply, and the question is no longer whether to act.
One question worth having ready. If a statin is offered and you are uncertain, do not ask whether you need it. Ask what it changes for you specifically, in absolute terms, over ten years. Any GP can produce that number, and it is far more informative than a relative risk reduction, which is identical for someone at high risk and someone at very low risk while meaning something completely different to each.
12. What we do not know
Whether steering by apoB produces better outcomes than steering by LDL. That apoB predicts better is settled[1]. That patients managed to an apoB target live longer than patients managed to an LDL target has not been trialled. A better predictor is not automatically a better instrument to treat by, and the distinction should be stated rather than glossed.
How much of the triglyceride benefit is real. Remove one trial and the estimate loses significance[7]. A meta-regression resting on a single influential result is a provisional answer.
Whether lowering Lp(a) helps. Causality is well supported. Benefit from lowering it is not yet demonstrated[9], and those are two different claims that get quoted as one.
What saturated fat should be replaced with. Polyunsaturated fat and carbohydrate both worked. Monounsaturated fat and protein had too little data to say[11]. Public dietary advice is more confident about this than the trials are.
Whether the training figures transfer to you. That meta-analysis studied sedentary people with metabolic syndrome. Somebody already fit has less headroom and should expect less[13].
And the caveat that applies to the whole page. A result is one morning. These values move between measurements with nothing behind it but biological and analytical variation. One number slightly over a line is an argument for a second sample, not for a conclusion.
13. Sources
- Marston NA, Giugliano RP, Melloni GEM, et al. Association of Apolipoprotein B-Containing Lipoproteins and Risk of Myocardial Infarction in Individuals With and Without Atherosclerosis: Distinguishing Between Particle Concentration, Type, and Content. JAMA Cardiology, 2022. PMID 34773460
- Sniderman AD, Navar AM, Thanassoulis G. Apolipoprotein B vs Low-Density Lipoprotein Cholesterol and Non-High-Density Lipoprotein Cholesterol as the Primary Measure of Apolipoprotein B Lipoprotein-Related Risk: The Debate Is Over. JAMA Cardiology, 2022. Editorial comment. PMID 34773457
- Ference BA, Ginsberg HN, Graham I, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. European Heart Journal, 2017. PMID 28444290
- Cholesterol Treatment Trialists' (CTT) Collaboration; Baigent C, Blackwell L, et al. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet, 2010. PMID 21067804
- Ference BA, Yoo W, Alesh I, et al. Effect of long-term exposure to lower low-density lipoprotein cholesterol beginning early in life on the risk of coronary heart disease: a Mendelian randomization analysis. Journal of the American College of Cardiology, 2012. PMID 23083789
- Nordestgaard BG, Varbo A. Triglycerides and cardiovascular disease. Lancet, 2014. PMID 25131982
- Marston NA, Giugliano RP, Im K, et al. Association Between Triglyceride Lowering and Reduction of Cardiovascular Risk Across Multiple Lipid-Lowering Therapeutic Classes: A Systematic Review and Meta-Regression Analysis of Randomized Controlled Trials. Circulation, 2019. PMID 31530008
- Voight BF, Peloso GM, Orho-Melander M, et al. Plasma HDL cholesterol and risk of myocardial infarction: a mendelian randomisation study. Lancet, 2012. PMID 22607825
- Kronenberg F, Mora S, Stroes ESG, et al. Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: a European Atherosclerosis Society consensus statement. European Heart Journal, 2022. PMID 36036785
- Nordestgaard BG, Langsted A, Mora S, et al. Fasting is not routinely required for determination of a lipid profile: clinical and laboratory implications including flagging at desirable concentration cut-points: a joint consensus statement from the European Atherosclerosis Society and European Federation of Clinical Chemistry and Laboratory Medicine. European Heart Journal, 2016. PMID 27122601
- Hooper L, Martin N, Jimoh OF, et al. Reduction in saturated fat intake for cardiovascular disease. Cochrane Database of Systematic Reviews, 2020. PMID 32827219
- Ho HV, Sievenpiper JL, Zurbau A, et al. The effect of oat β-glucan on LDL-cholesterol, non-HDL-cholesterol and apoB for CVD risk reduction: a systematic review and meta-analysis of randomised-controlled trials. British Journal of Nutrition, 2016. PMID 27724985
- Wood G, Taylor E, Ng V, et al. Determining the effect size of aerobic exercise training on the standard lipid profile in sedentary adults with three or more metabolic syndrome factors: a systematic review and meta-analysis of randomised controlled trials. British Journal of Sports Medicine, 2021. PMID 34193471
More free knowledge at kratosnatural.com.
© Kratos Natural. This guide is free: read it, keep it, print it, give it to whoever you like. What you may not do: sell it, or present it as your own work. The knowledge belongs to everyone. The text stays ours.