Article at a Glance
- Lipoprotein(a), also called Lp(a), is a lipoprotein subclass that has been linked to increased risk of heart disease when elevated.
- Although elevated Lp(a) numbers are thought to be tied to genetics and tough to lower as a result, dietary changes and supplements like Niacin, fish oil, and L-carnitine can work for some people.
Lipoprotein(a), also called Lp(a), has gotten a good deal of press lately as a metric to watch if you’re concerned with heart health. Elevated Lp(a) levels have been linked with an increased risk of cardiovascular disease, with the Framingham study reporting a 2 to 3 times increased risk for heart disease when Lp(a) levels are greater than 30mg/dl, or 75 nmol/L. (R)
I was surprised to see the findings of this JAMA study which equated elevated Lp(a) with a risk for heart disease similar to having total cholesterol of 240 mg/dl, even in men under 55 years of age. However, we have seen very public examples of how elevated Lp(a) can do damage early in life. Celebrity trainer Bob Harper, of Biggest Loser fame, suffered a heart attack at age 52 when he appeared to be in the prime of his life. As it turns out, the culprit was elevated Lp(a). Bob’s doctors hadn’t tested him and so he never knew he was at increased risk for heart disease.
Even had he known could Bob have lowered Lp(a) enough to avoid a heart attack? The subject is a matter of some controversy. Unlike other cardio related biomarkers, Lp(a) numbers are reputed to be tough to move with diet and drugs.
- Is it possible to lower Lp(a) with diet?
- John's Lp(a) progress
- Lipoproteins – cholesterol taxis
- LDL vs. HDL
- LDL-P vs. LDL-C
- Lp(a)’s relation to LDL-C
- Measuring Lp(a) – mg/dl vs. nmol/L
- Most people aren’t tested for Lp(a)
- How I lowered my Lp(a)
- Supplements that lower Lp(a)
- A note on vegetable oil and Lp(a)
- Lp(a) – it’s all in the gene?
- Additional resources
Is it possible to lower Lp(a) with diet?
At least this is what I was told by my doctor when a blood test showed I had Lp(a) markers that were relatively high at 49mg/dl.
“Lp(a) is genetic, you can’t change it all that much. It’s possible Niacin could help”
That’s never fun to hear for a metric that is tied to increased risk of heart disease, but I resolved to try nonetheless.
As I outline later in this post, my first course of action was to have a retest as I believed at the time that mg/dl wasn’t an accurate way to measure Lp(a). I have since learned that I was wrong. Many health commentators on the podcast circuit will tell you that a weight based measurement like mg/dl can register a high Lp(a) number based on large fluffy particles, which “tip the scales” and which aren’t particularly atherogenic. Instead of mg/dl, they advocate for mol/L. However, both mg/dl and nmol/L are mass based measurements, and in my case, they tracked very closely (as they will for everyone else since they mutually convert). What commentators like Chris Kresser, who is one of the ones wrongly calling for nmol/L measurements, are really trying to say is that an Lp(a) particle count is preferable to either mg/dl or mol/L since these mass based measurements could come back on the high side if the particles are large and fluffy. Lp(a) particle gives you the total number of Lp(a) particles with the potential to do damage and is therefore a greater predictor of risk, although as a practical matter, lipid experts like Tom Dayspring say that the weight based Lp(a) measurements will almost always track with the particle count. As of writing this update, I have yet to have my Lp(a) particle count done.
After the unnecessary retest, my next move was to experiment with diet. As you can see from the chart below, I was able to drop my Lp(a) number down to 33mg/dl by focusing on lowering my LDL-C and LDL-P numbers.
John's Lp(a) progress
So, we begin this post with a teaser, and an optimistic one at that: it is absolutely possible to lower your Lp(a) levels with diet and lifestyle changes.
As you move through this article, I dive more into my own experiences testing my Lp(a) levels, what supplements have shown promise in reducing Lp(a), how Lp(a) is tied to LDL, and how I was ultimately successful in lowering my numbers.
Towards the end of the post Aaron weighs in on the genetics of Lp(a).
We begin with a refresher on Lipoproteins, because if you don’t understand the basics of lipoproteins, you can’t have a good understanding of Lp(a).
Lipoproteins – cholesterol taxis
For years, dietary cholesterol has been demonized as something to avoid, but the fact is, we need some cholesterol in our diet. We need cholesterol to use vitamin D, build essential hormones, and for our brains to function properly. (R) Furthermore, most of the cholesterol in the body is made by the body, it doesn’t come from food at all. (R) Having said that, people absorb different amounts of the cholesterol they eat. One place to look to see whether you are a hyper absorber of cholesterol are the ABCG8 genes. ABCG8 is responsible for pushing out cholesterol and plant sterol that makes its way past the gut wall (the lumen) back into the digestive track where it can be excreted in bile. Thing is, the amount of cholesterol and sterol people absorb varies greatly. If you’re someone who absorbs higher amounts of cholesterol, this can contribute to a spike in LDL-C, which then makes elevated Lp(a) more dangerous according to some clinicians.
But again, dietary cholesterol itself is not necessarily the enemy, it’s how that cholesterol is transported through the body by special proteins known as lipoproteins. Cholesterol is only a problem when it clogs an artery, and it is certain types lipoproteins that take these fats and deposit them at the artery wall.
Think of lipoproteins as the yellow taxis that shuttle cholesterol around the body. They literally encase fat, just as a taxi does for a human, and deliver it to different regions of the body.
Lipoproteins have an important role to play in how the body transports fat, but not all lipoproteins are created equal. Just as there are different breeds of dog, there are different types of lipoprotein.
The two big ones are LDL and HDL.
LDL vs. HDL
HDL stands for “high density lipoprotein.” HDL sends cholesterol to your liver where it is metabolized and excreted. (R) HDL is generally labeled as good since it acts to remove unnecessary cholesterol from the body. apoA molecules are the proteins associated with HDL.
By contrast, LDL, or “low density lipoproteins” when they take a “wrong turn” can cause fat build up within the walls of your arteries, and is thought to increase the risk for heart attack and stroke. (R) (R) apoB molecules are the proteins associated with LDL.
LDL-P vs. LDL-C
It is important to know the difference between LDL-C and LDL-P if you’re going to understand Lp(a).
LDL-C, long the target of statin therapy, represents the amount of cholesterol carried by LDL in your body. You can have low LDL-C, and still have a high risk of heart disease because of LDL-P, or the total number of LDL particles in your blood.
LDL doesn’t just carry cholesterol, it also carries triglycerides, another form of fat that generally rises with sugar consumption. When blood work comes back and triglycerides are high, this can mean “cholesterol depleted” LDL, which results in low LDL-C, but elevated LDL-P. In other words, the LDL particle represents the total number of taxis on the road, not just the taxis carrying cholesterol.
It is now thought that elevated LDL-C is only a predictor of heart disease when it is concordant with high LDL-P. Low LDL-C and high LDL-P is the most dangerous territory for heart disease. (R)
With that out of the way, let’s delve into Lp(a), a great topic for our blog because Lp(a) is thought to be driven largely by genetics.
Lp(a)’s relation to LDL-C
Lp(a) is a type of lipoprotein related to LDL. In structure, Lp(a) is a cholesterol rich LDL particle with an extra Apo(a) protein added on, hence the “little a.” (R)
While it is generally agreed that high Lp(a) levels are not a good thing, the exact levels considered safe, as well as how to measure for Lp(a) is a subject of some controversy.
Lp(a) first came on my radar after doing a Boston Heart Diagnostics Cardio panel. My lipid markers were generally ok, but at 43 mg/dl, at my last test, and with previous results at 46 and 49 mg/dl, my Lp(a) number was labeled borderline.
My doctor began discussing Niacin as a supplement, as there is some data to suggest that Niacin can effectively reduce Lp(a) levels, although this study suggests that Niacin’s effectiveness may vary based on Apo(a) phenotype. (R)
We will get to Niacin and it’s ability to lower Lp(a) in more detail later on.
Measuring Lp(a) – mg/dl vs. nmol/L
As I dug into the Lp(a) research, I initially believed there was a distinction between testing for Lp(a) with mg/dl vs. nmol/L. I believed that mg/dl was a weight based measurement and that nmol/L was concentration based, and that mg/dl could be thrown off by large buoyant particles, whereas nmol/L would give you a more accurate concentration based measurement.
Some commentators argue (wrongly) that mg/dl is not an accurate way of testing for Lp(a) since the large, buoyant particles can throw off the weight and it’s the small dense particles that cause heart disease. It’s true that the small dense particles are the culprits, but the issuer is not with mg/dl vs. nmol/L. In fact, nmol/L converts to mg/dl, so both are valid. As I discussed above, the Lp(a) particle count would be the ideal and most accurate way of determining risk.
Differences in Lp(a) reporting
Interestingly, Boston Heart Labs lists Lp(a) results on its Cardio panel in mg/dl, while the Quest Diagnostics Cardio IQ panel lists Lp(a) results in nmol/L. Again, this was an unnecessary step, but I had a Boston Heart panel done, and then followed up a week later with a Cardio IQ panel from Quest. I was looking for a disparity between the mg/dl number and the nmol/L number to try to prove, or disprove, the theory that nmol/L is the most accurate way to measure for Lp(a) (again the mg/dl vs. nmol/L is not an accurate theory).
Unsurprisingly, my labs came back with an Lp(a) of 43 mg/dl and 98 with a nmol/L measurement, in the borderline category on both tests.
The European Atherosclerosis Society (EAS) lists Lp(a) numbers of >50 mg/dl as elevated, so my numbers wouldn’t have been considered elevated by that group, however, they are considered on the high side by most experts in the field in the US. (R)
Most people aren’t tested for Lp(a)
Most “normal” lipid panels don’t include Lp(a).
Instead, Lp(a) “hides” in the LDL-C number. What I mean here is that, absent a break out for Lp(a) specifically, Lp(a) will be included in your overall LDL-C number. (R) This is an issue because it could indicate statin therapy for those who can’t reap the benefits, as Lp(a) doesn’t respond to statin therapy.
For more on this theory, see this study from Kidney International.
Since statins have no influence on Lp(a) levels, it can be expected that the LDL cholesterol-lowering effect of statins may be diminished in patients who have a pronounced elevation of Lp(a) levels accompanied by only moderate elevations of LDL cholesterol.
When LDL numbers were adjusted for Lp(a), 25.7% of the patients in the study had LDL levels that were no longer in need of statin intervention.
How I lowered my Lp(a)
When I first learned that I had elevated Lp(a), I was concerned. I was especially concerned because many say there isn’t much you can do to lower Lp(a).
Having said that, since I first found out about elevated Lp(a) levels, I’ve moved my number down 16 points (33%), from a high of 49 mg/dl to my new low of 33 mg/dl. Again, this is not scientific, but between blood draws, which I do every 3-4 months, I experiment with different nutritional strategies.
I achieved my previous low Lp(a) number (43) after shifting to a largely plant based diet, completely getting rid of added butter (when I go out to eat all bets are off if butter is included in a dish), eating egg whites instead of the whole egg, and eating red meat very sparingly, if at all. I already eat clean, so there’s no telling whether these changes contributed to the decline of my Lp(a) number.
I achieved my 33mg/dl number after transitioning to a 85-90% plant based diet. So, essentially, I ignored most of the Bulletproof ketogenic diet themes and lowered the amount of animal protein and animal fat I was consuming. I ate Vegan many days a week and almost always had a plant based breakfast and lunch. In short, I ate a lot less meat and a lot less saturated fat. I would still eat eggs once or twice a week, wild salmon, and chicken sparingly, but I became a vegetarian for the most part.
I am not suggesting this will work for everyone, it’s just what worked for me.
Lower LDL means lower Lp(a)
There are studies that offer good news for those of us who need to keep an eye on Lp(a). Since Lp(a) is carried on the LDL particle, it needs LDL to do damage. The less LDL to bind to, the less Lp(a). (R)
The Cleveland Clinic references a study on this blog page, without citing or linking to it, that looked at 5,000 patients with elevated Lp(a). When the LDL in these patients was brought down (not specified whether this was LDL-C or LDL-P), the “increased risk for mortality from Lp(a) was negligible.”
Strategies for lowering LDL-P
This does not apply to everyone, I fully get that, but the best method I have found for lowering LDL-C and Lp(a) has been eating a more plant based diet. This does not mean an exclusively plant based diet, but it does mean multiple Vegan meals a week. That dietary intervention is where I saw my greatest results thus far.
If the goal for those of us with elevated Lp(a) is to lower LDL-P count, there is a fork in the road based on whether you are insulin resistant, i.e. whether you have elevated triglycerides and blood sugar numbers. If the answer is yes, it may be better to focus on a diet that is lower in high glycemic foods.
This study found that carbohydrate restriction reduced LDL-P in overweight men.
The idea here is that, in insulin resistant patients, the LDL particle is packing greater amounts of triglycerides (produced from sugar) than cholesterol. This puts the LDL-C number low, but the LDL-P, the true predictor of bad heart outcomes, high.
Taking care of the insulin resistance with metformin, reduces the LDL-P count, i.e. the number of LDL-P that are moving triglycerides around. This is a perfectly logical result in those with high triglycerides (as was the case with the case study) and high LDL-P, but not as much in cases where LDL-P is elevated and triglycerides are low.
In those cases, where the LDL particles are cholesterol rich and metabolic syndrome can be effectively ruled out, we can’t rely on a “cholesterol depleted” lipoprotein theory.
In cases where the LDL-C number is high, and triglycerides are low, the method of attack is more likely to be a diet that is very low in fat and cholesterol. For more on lowering LDL-C, check out this post: Why is my LDL-C high and what can I do to lower it?
Supplements that lower Lp(a)
Lysine and vitamin C
Also known as the Linus Pauling Lp(a) protocol, this regimen calls for 3g of Vitamin C and 3g of Lysine a day.
There is some chatter indicating that Lysine and vitamin C supplementation can reduce Lp(a) numbers because lysine binds to Lp(a) particles. The lysine doses recommended sound very high to me, and I would imagine they would come with side effects for many people.
Niacin and lowering Lp(a)
The bottom line here is twofold:
- Niacin supplementation has been shown to lower Lp(a) levels significantly (R) (R) (R) (R)
- Niacin has side effects such as flushing and insomnia (R)
Omega 3 fish oil and Lp(a)
This one surprised me. As you can see from the table above, there are a series of studies that show omega 3 fish oil, sometimes alone and sometimes combined with other supplements and diet changes, lowers Lp(a).
If you’re in the market for fish oil products, check out our guide to Omega 3 fish oil, which I only half jokingly titled “Most Fish Oil is Garbage. Here’s What to do About it.”
A note on vegetable oil and Lp(a)
Let’s revisit for a second the issue of cholesterol absorption and the ABCG8 genes I touched on at the outset of the post. Perhaps the reason I had success lowering my Lp(a) by targeting LDL-C is that I am a hyper absorber of cholesterol. In some people, the cholesterol they eat gets kicked out of the lumen by the ABCG8 genes and therefore never makes its way into the blood. In my case, I have some ABCG8 SNPs as well as elevated levels of sitosterol to prove that I am absorbing more cholesterol and sterol in general. For those of you who aren’t familiar, many lipid panels measure sterol absorption as a proxy for cholesterol absorption. Since the body makes most of its own cholesterol, we can’t use cholesterol as a marker for cholesterol absorption because it’s already there, so we look at markers like sitosterol, which is a plant fat found in foods like oats, avocado, beans, and yes, in vegetable oils. My sitosterol levels have been as high as 3.9 mg/dl, which isn’t in sitosterolemia territory, but is evidence for greater absorption. It’s generally regarded as a bad thing to be absorbing plant sterols as they are only thought to be heart healthy to the extent they prevent the absorption of cholesterol and then are kicked back out of the lumen. When they get absorbed, especially in oxidized form, as they appear in vegetable oils (which are notoriously unstable fats), they can do real damage, especially for those of us with elevated Lp(a). (R)
This NEJM study blew my mind. It teaches us that oxidized phospholipids (like we find in abundance in vegetable oils and other processed plant fats) preferentially bind to Lp(a), and it’s the combination of the oxidized phospholipids married to the Lp(a) that makes them especially dangerous.
Lp(a) – it’s all in the gene?
Why are Lp(a) numbers though to be relatively static? In part because they’re linked to genetics.
If you have elevated Lp(a), you likely inherited it, which means it’s that much more difficult to transform with diet. To explain the genetic side of Lp(a), I’m bringing in Aaron to do his thing.
Passing the mic to Aaron for this section.
Thanks John. Lp(a) production is mainly controlled by the LPA gene. The LPA gene is one of those interesting genes which contains something called a variable number tandem repeat region (VNTR), much like MAOA which I’ve talked about previously. VNTRs are regions where a simple sequence of DNA can be repeated multiple times, with the number of repeats varying between individuals.
In the case of LPA, the number of repeats can vary between 10 and 50. In those with a larger number of repeats, Lp(a) produced from LPA is many times larger. (R) Interestingly, there is a very strong negative-correlation between Lp(a) size and its concentration in the blood (R); i.e. the more repeats in your LPA gene, the larger your Lp(a) protein and the lower your Lp(a) concentration in the blood.
Or to flip it around the other way, with fewer repeats in your LPA gene you will have a smaller Lp(a) protein but a higher concentration in the blood.
The exact reason why a bigger protein leads to a reduced concentration in the blood isn’t known, but scientists have hypothesized quite simply that a larger protein takes longer to make, and so there will be less of it produced (R).
LPA and SNPs?
There are two SNPs within LPA which are associated with an increased cardiovascular risk: rs3798220 and rs10455872.
In a similar way to MAOA neither are thought to change protein activity, rather they are thought to associate with VNTR numbers. rs3798220 or A5673G describes the switch from a ‘T’ allele to the risk ‘C‘ allele in approximately 5% of the population. Those carrying the ‘C‘ allele had increased blood Lp(a) and had double the cardiovascular risk.
Similarly, rs10455872 or T3947+467C, which describes the switch from an ‘A’ allele to a risk ‘G‘ allele was also associated with increased blood Lp(a) and an increased cardiovascular risk, although to a lesser extent than rs3798220 A5673G.
I hope that helps with your understanding of the genetics of Lp(a), as always happy to discuss any queries in the comments.
In conclusion, just as with nutrigenomics, where we cannot hyper focus on one SNP, there must be nuance as we examine our Lp(a) numbers as well. We know that Lp(a) is driven by genetics, but Lp(a) must bind with an LDL particle in order to do harm. For this reason, those facing elevated Lp(a) are well served in understanding what drives LDL, especially LDL-P.
In cases where triglycerides are high, it appears insulin sensitivity may be the primary driver of LDL-P. Absent insulin sensitivity, you have to really put your detective hat on to hack your own system and figure out how best to keep your LDL-P at a level where Lp(a) damage is mitigated.
For a different perspective on LDL, see LDL is Your Friend, by Dr. David Perlmutter (he’s referring to LDL-C).
For an epic explanation of how our body’s use cholesterol, and which metrics actually matter for heart disease, I highly recommend Dr. Peter Attia’s lengthy, but excellent, series: the Straight Dope on Cholesterol.
Aaron also put together a valuable summation of the interplay between cholesterol, heart health, and genetics on this Gene Food blog, which is worth a read.