I notice in my everyday practice and in social media that in the “cholesterol skepticism” community -often overlapping with parts of the low-carb world- the spotlight has recently shifted to lipoprotein(a), or Lp(a). And not just shifted in a neutral way. The language has become explicitly causal: “LDL is not causal, but Lp(a) is.”
If that is now the argument, then the conversation is, in a sense, already over — because the claim collapses under its own logic. Here is why.
How science decides whether something is causal
Researchers do not decide causality based on preference, ideology, or online narratives. They rely on established criteria, often summarized through the Bradford Hill framework: strength of association, consistency, temporality, dose-response relationship, biological plausibility, reversibility, experimental evidence, and more. LDL cholesterol meets these criteria convincingly.
There are more than 200 large studies showing a clear dose-response relationship between LDL levels and cardiovascular risk. Randomized clinical trials, taken together, include roughly 2 million participants and more than 150,000 cardiovascular events. And these trials consistently show that lowering LDL reduces risk. Studies have also shown that bringing LDL below 70 mg/dL can lead to regression of atherosclerosis. On top of that, Mendelian randomization studies demonstrate that people who are genetically exposed to lower LDL levels from birth have a dramatically lower risk of coronary heart disease.
In other words, the evidence for LDL as a causal factor in atherosclerotic cardiovascular disease is not marginal. It is one of the strongest causal cases in preventive cardiology.
What about Lp(a)?
Lp(a) also fulfills many of the same criteria pointing toward causality. Epidemiology, dose-response data, biological plausibility, and Mendelian randomization studies all support the idea that Lp(a) is not merely associated with cardiovascular disease, but actively involved in its development.
What is still missing is the final piece: a clinical trial specifically designed to test whether directly lowering Lp(a) reduces cardiovascular events.
There are already encouraging signals. In analyses from PCSK9 inhibitor trials such as evolocumab, a 27% reduction in Lp(a) was associated with a 23% reduction in cardiovascular risk. But this still does not count as definitive proof, because those drugs also lower LDL substantially. The dedicated outcome trial meant to answer this question more directly is Lp(a)HORIZON, a large Phase III study involving around 8,000 participants. Its purpose is straightforward: to test whether specifically lowering Lp(a) translates into fewer cardiovascular events.
If the results are positive, Lp(a) will stand beside LDL as another established causal risk factor.
The interesting point you should know
Lp(a) contains an LDL particle. Literally. The “a” in Lp(a) refers to apolipoprotein(a), a protein attached to what is otherwise a standard LDL-like particle. You can think of it as an LDL particle with an added feature that makes it more adhesive, more biologically disruptive, and more likely to contribute to plaque formation. So when someone says, “Lp(a) is causal but LDL is not,” they are making a claim that is internally contradictory. They are implicitly admitting that cholesterol-carrying lipoprotein particles can cause cardiovascular disease when they enter and are retained in the arterial wall. But that is precisely the core mechanism long established for LDL as well.
The deeper point
The issue becomes even clearer when you look at comparative genetic data. A study by Burgess and colleagues found that approximately a 100 mg/dL reduction in Lp(a) is needed to produce a similar reduction in coronary disease risk as a 38 mg/dL reduction in LDL cholesterol. What matters here is not just the raw mass difference, but what those particles are carrying. The cholesterol content represented by those changes is broadly comparable, because both Lp(a) and LDL transport cholesterol into the arterial wall.
That is the key point: regardless of whether cholesterol is being delivered by LDL or by Lp(a), cholesterol deposition within the arterial wall is the process that drives atherosclerosis. Lp(a) may add extra danger through inflammatory, prothrombotic, and adhesive properties. But it does not replace the LDL mechanism. It reinforces it.
The right question for your heart
When someone argues that Lp(a) causes cardiovascular disease but LDL does not, they are effectively conceding the very principle they are trying to deny: that cholesterol-carrying lipoprotein particles, once they penetrate and become trapped in the arterial wall, initiate and promote atherosclerosis.
Science does not work by forcing us to choose one “true” cause while dismissing all others. Multiple factors can be causal at the same time. LDL is causal and Lp(a) is very likely causal as well. And the existence of one strengthens the biological case for the other rather than weakening it.
The same logic applies to the broader discussion around saturated fat. The real issue is not choosing tribes or defending dietary identities. The real issue is mechanism and clinical outcomes. It is one thing to say that in some populations the incremental benefit of further reduction may be smaller and it is something entirely different to claim that saturated fat is harmless, or that higher intake is therefore a good idea.
The relevant question is always the same: what replaces it, and what is the net effect on risk and benefit for the actual person in front of you? Polyunsaturated fats are not the same thing as refined carbohydrates. A whole dietary pattern is not the same thing as a social media slogan.
That is where the conversation should be.
Resources
1. Ference et al. (2012) — JACC
“Effect of Long-Term Exposure to Lower LDL-C Beginning Early in Life on the Risk of CHD: A Mendelian Randomization Analysis”
DOI: 10.1016/j.jacc.2012.09.017
2. Burgess et al. (2018) — JAMA Cardiology
“Association of LPA Variants With Risk of Coronary Disease and the Implications for Lipoprotein(a)-Lowering Therapies”
DOI: 10.1001/jamacardio.2018.1470
3. O’Donoghue et al. (2019) — Circulation
“Lipoprotein(a), PCSK9 Inhibition, and Cardiovascular Risk: Insights From the FOURIER Trial”
DOI: 10.1161/CIRCULATIONAHA.118.037184 |
4. Ference (2015) — Current Opinion in Lipidology
“Mendelian Randomization Studies: Using Naturally Randomized Genetic Data to Fill Evidence Gaps”
DOI: 10.1097/MOL.0000000000000247 |
5. O’Donoghue et al. (2022) — Circulation
“Long-Term Evolocumab in Patients With Established ASCVD (FOURIER-OLE)”
DOI: 10.1161/CIRCULATIONAHA.122.061620 | 238 citations
6. Gaba et al. (2023) — Circulation
“Association Between Achieved LDL-C Levels and Long-Term Cardiovascular and Safety Outcomes: FOURIER-OLE”
DOI: 10.1161/CIRCULATIONAHA.122.063399 | 108 citations
7. Cohen et al. (2006) — NEJM
“Sequence Variations in PCSK9, Low LDL, and Protection Against Coronary Heart Disease”
DOI: 10.1056/NEJMoa054013 | 3,032 citations
8. CTT Collaboration (2010) — Lancet
“Efficacy and Safety of More Intensive Lowering of LDL Cholesterol”
DOI: 10.1016/S0140-6736(10)61350-5
