Pairing Genetic Risk With Blood Biomarkers in Cardiovascular Prevention

Drew, I'm seeing this more and more — patient walks in with a consumer genomics report, top decile for coronary artery disease polygenic risk score, lipid panel flagged normal. She wants to know if she should be worried.

The short answer is that her lipid panel answered the wrong question. A polygenic risk score — a number summing the small effects of hundreds of common variants into a single probability estimate — tells you the starting hand. It never changes. The blood tests tell you how that hand is currently being played.

So the LDL-C of 118, which the lab flagged normal, isn't actually reassuring?

Not by itself. LDL-C is a calculated estimate and it systematically undercounts atherogenic particle burden. ApoB — apolipoprotein B, one molecule per atherogenic particle — is guideline-endorsed by the European Society of Cardiology as the superior target. ESC 2021 says high-risk patients should be below 65 mg/dL. That's an approved, guideline-backed threshold.

What's a realistic apoB in someone with her LDL?

Anywhere from 80 to over 110. They don't track tightly, especially with elevated triglycerides or insulin resistance. That's why you order it directly. It's a standard lipid add-on, not exotic.

Her report also flagged an LPA variant. What do I do with that?

Order a serum Lp(a) — lipoprotein(a), a genetically determined atherogenic particle. The genotype predicts high levels but doesn't give you the number. ESC recommends measuring it once in every adult. Above 125 nmol/L is high-risk. There's currently no approved treatment that reduces Lp(a) endpoints, though RNA-based trials are ongoing.

So I measure it, I tell her the number, and then what — I watch?

You watch, and you treat everything else more aggressively. High Lp(a) shifts the entire risk calculation. It's also an argument for earlier coronary calcium scoring if she's borderline on statin initiation. That's guideline-consistent reasoning, not speculation.

Her fasting glucose was 94. That's technically normal. Should I be looking harder?

Yes. Fasting glucose is the last thing to rise in insulin resistance. Fasting insulin alongside it gives you HOMA-IR — a calculated index, not a reference-range-validated diagnostic, but with strong mechanistic and epidemiologic support. Triglyceride-to-HDL ratio above roughly 1.5 in mg/dL units is a validated surrogate for insulin resistance with outcomes data. Her ratio here is under 1.7 — marginal, worth tracking.

How strong is the HOMA-IR evidence for actually changing management?

Mechanistic and observational. No randomized trial has used HOMA-IR as an enrollment criterion and shown a hard endpoint benefit from treating to a HOMA-IR target. I order it to identify the pattern early and motivate lifestyle work — not as a drug-initiation threshold.

She's also heterozygous for HFE C282Y — the hemochromatosis variant. Does that interact?

Heterozygosity alone rarely causes iron overload. Check a fasting transferrin saturation and ferritin. If normal, reassure and recheck in a few years. If ferritin is elevated, refer to hepatology. Don't treat the genotype — treat the phenotype.

What's the one thing you'd want her primary physician to hold onto from all of this?

A laboratory reference range tells you what's common in the population that was tested. It is not a target. ApoB below 65, Lp(a) measured once, insulin resistance pattern caught early — those are targets. The genomics report told you where to look. The blood tests tell you whether the risk is being expressed right now.