The Lp(a) risk levels
| Lp(a), nmol/L | Lp(a), mg/dL | Risk level | Where the line comes from |
|---|---|---|---|
| Under 75 | Under 30 | Low risk | Guideline line. National Lipid Association (2024). |
| 75 to under 125 | 30 to under 50 | Intermediate risk | Guideline band. National Lipid Association (2024). |
| 125 or higher | 50 or higher | High risk: talk to your doctor | Guideline line. National Lipid Association (2024); ACC/AHA 2018 and 2026. Bring it to your next appointment. It is not an emergency. |
Phoenix's choice here is the framing, not the numbers. Phoenix shows Lp(a) as your once-measured risk level, with no "optimal" target, because you cannot work it down.
The line rule. "Under 75" excludes 75, and "125 or higher" includes 125. So 75 nmol/L is intermediate and 125 nmol/L is high.
Use the unit your lab printed. nmol/L and mg/dL do not convert exactly (see "How do you get Lp(a) tested?").
Confidence in the categories: high. They are the lipid society's own categories. None was built from carrier data.
Why is Lp(a) a risk level and not a target?
Because your genes set it. Lp(a) levels are genetically determined (Tsimikas, 2020), about 1 in 5 people have a raised level (Koschinsky, 2024), and no approved medicine is built to lower it yet (Bene-Alhasan, 2026).
It still matters. Large genetic and population studies show Lp(a) causes heart disease and narrowing of the aortic valve, and raises risk even when LDL cholesterol is very low (Kronenberg, 2022).
So a high Lp(a) is information, not a failure. It tells you and your doctor how hard to push on the things you can move.
Does Lp(a) matter for the brain or for APOE4?
Not for Alzheimer's across the normal range. In 539,478 people followed for up to 30 years, Lp(a) levels were not linked to Alzheimer's disease or vascular dementia (Thomas, 2025). One genetic marker of small Lp(a) particles, the kind that go with the highest levels, did go with a 25% higher Alzheimer's rate (hazard ratio 1.25), so very high levels are not in the clear.
APOE4 and Lp(a) are separate stories. Lp(a) is set by a different gene, LPA, and no useful study links the two (the only one is a 124-person case-control study; Solfrizzi, 2002). For a carrier, Lp(a) is a heart and valve marker. Test it because it changes how strict your ApoB and LDL-C goals should be.
What do the guidelines say?
- National Lipid Association (2024): measure Lp(a) at least once in every adult. Under 75 nmol/L (30 mg/dL) is low risk, 75 to under 125 (30 to under 50) intermediate, and 125 nmol/L (50 mg/dL) or more high. Testing the parents, siblings and children of anyone with a high level finds more people who need care. People with a high level get earlier, more intensive risk-factor care, mainly lowering LDL-C (Koschinsky, 2024).
- European Atherosclerosis Society (2022): test at least once in adults, and manage overall risk early and intensively (Kronenberg, 2022).
- ACC/AHA (2018): 50 mg/dL or 125 nmol/L or more is a risk-enhancing factor (Grundy, 2018).
- ACC/AHA (2026): measure once in adulthood; 125 nmol/L or 50 mg/dL or more is high.
What did Phoenix choose, and why?
I chose to show Lp(a) as a risk level, using the lipid society's three categories, with no Phoenix target.
A target would ask you to move a number your genes fixed. That sets people up to feel they failed. A risk level tells the truth and points to the work that pays.
That work is ApoB and LDL-C. In the FOURIER trial, people with Lp(a) above the median got a 23% drop in coronary events from LDL lowering with evolocumab, against 7% for people below it (O'Donoghue, 2019). If your Lp(a) is high, Phoenix's ApoB line of under 60 matters even more for you.
How do you get Lp(a) tested?
Ask for "lipoprotein(a)" by name. Most standard lipid panels leave it out, and testing rates are low even though guidelines now ask for it once in every adult (Koschinsky, 2024).
Labs report it in two units. nmol/L counts particles. mg/dL weighs them. The particles come in different sizes from person to person, set by how many repeats your LPA gene carries (Thomas, 2025), so there is no single conversion factor. Use the category in the unit on your report: 75 nmol/L sits with 30 mg/dL, and 125 nmol/L with 50 mg/dL.
What moves Lp(a)?
Very little that is available today. Ranked by size of effect:
- New Lp(a) drugs in trials: 80% to almost 100% lower. Olpasiran cut Lp(a) by 97% or more at its higher doses (O'Donoghue, 2022). Pelacarsen cut it by up to 80% (Tsimikas, 2020). Neither is approved. Outcome trials are running to show whether the drop prevents heart attacks (Bene-Alhasan, 2026), and the pelacarsen trial, Lp(a)HORIZON, is due to report in late 2026 (Ebubechukwu, 2026).
- Lipoprotein apheresis. A blood-filtering procedure. Its US approval for Lp(a) covers people with familial hypercholesterolemia and artery disease whose Lp(a) stays at 60 mg/dL (about 150 nmol/L) or more and LDL-C at 100 mg/dL or more on full treatment (Koschinsky, 2024).
- PCSK9 inhibitors: about 27% lower. Evolocumab cut Lp(a) by a median of 26.9% (O'Donoghue, 2019).
- Oral hormone therapy after menopause: about 20% lower. Across 24 studies, hormone therapy lowered Lp(a) by 20% against placebo or no treatment, and oral estrogen lowered it more than patches or gels (Anagnostis, 2017).
- Statins: 8 to 20% higher. Statins raise Lp(a) a little: a mean rise of 8.5 to 19.6% across six trials (Tsimikas, 2020). They still cut heart events through LDL, so this is not a reason to stop one.
- Diet and exercise. Neither guideline lists a diet or exercise program that lowers Lp(a). Their advice for a high level is to lower everything else (Koschinsky, 2024; Kronenberg, 2022).
Among Phoenix members (as of 2 October 2026), 80 have logged a statin and 14 a PCSK9 inhibitor or similar injection, so the small statin rise applies to far more members than the PCSK9 drop.
How often should you retest?
Once is enough for most adults. The National Lipid Association, the European Atherosclerosis Society and the 2026 ACC/AHA guideline all say to measure Lp(a) at least once (Koschinsky, 2024; Kronenberg, 2022).
Phoenix members' results back this up. Of 32 Phoenix members who uploaded two Lp(a) results in nmol/L, 28 (88%) stayed in the same risk category, and the median result changed by 1.2% over a median of 204 days.
My rule of thumb: test again only when something known to move it has changed, such as starting hormone therapy (about 20% lower), a statin (8 to 20% higher) or a PCSK9 inhibitor (about 27% lower), or when your first result sat right on 75 or 125.
My numbers
I tested Lp(a) once: 2.9 mg/dL on 9 May 2025, well inside the low-risk category (under 30 mg/dL). I have not retested, and the guidelines say I do not need to.
I am one of the lucky ones, and I only know because I asked for the test. This is one carrier's result, and yours can be very different: APOE4 and Lp(a) are separate genes.
What Phoenix members' results show
Carriers are not spared a high Lp(a). Of 122 Phoenix members who uploaded an Lp(a) result in nmol/L (as of 2 October 2026), 93 (76%) are in the low-risk category, 16 (13%) are intermediate, and 13 (11%) are at 125 nmol/L or higher. Among the 67 members with two copies of APOE4, 49 (73%) are low risk and 10 (15%) high. So about 1 in 4 members tested sits above the low-risk line.
Sixteen more members have results in mg/dL. That is too few to split, and Phoenix never converts them into nmol/L to pool them.
When should you talk to your doctor?
Talk to your doctor if your Lp(a) is 125 nmol/L (50 mg/dL) or higher. That is the high-risk line in the lipid society's categories and a risk-enhancing factor in the ACC/AHA guideline.
Three questions are worth taking with you:
- Given my Lp(a), how low should my ApoB and LDL-C be?
- Should my parents, siblings and children be tested? (The National Lipid Association backs this family testing.)
- Is there a clinical trial of a new Lp(a) drug that fits me?
Between 75 and under 125 nmol/L, mention it at your next appointment alongside your ApoB and LDL-C. If you take a statin, keep it exactly as prescribed; its small effect on Lp(a) is not a reason to change it on your own.