You may have heard that half of all heart attacks happen in people with normal cholesterol. While a popular line, and it’s wrong. A 2025 JACC study of 9.3 million adults found that 99.7% of heart attacks were preceded by at least one abnormal traditional biomarker. Stroke, heart failure, and total cardiovascular disease showed nearly identical numbers.

The source of the discrepency is that you have to look back in time. Lab values measured the day someone has a heart attack are misleading since, for example, people on a statin look like they have normal cholesterol but will have had a many-year history of elevated cholesterol. Furthermore, the thresholds used to label a value “abnormal” were often set at the disease level (cholesterol of 240, blood pressure of 140/90), not the level where heart attack risk starts to climb.

This post covers the 2025 JACC study, why earlier research undercounted the true predictability of heart attacks, and how the six core heart biomarkers (ApoB, Lp(a), hs-CRP, blood pressure, HbA1c, and eGFR) fit into the picture.

Source: Lee et al., JACC 2025

99.7% of heart attacks are preceded by an abnormal biomarker

The 2025 JACC paper by Lee and colleagues followed two large cohorts: KNHIS (9,341,100 Korean adults) and MESA (6,803 American adults). Across both cohorts, the authors checked whether each person had any of four traditional risk factors at any point before their first cardiovascular event:

Risk factor	Threshold
Blood pressure	SBP ≥120 mm Hg, DBP ≥80 mm Hg, or on antihypertensive treatment
Cholesterol	Total cholesterol ≥200 mg/dL, or on lipid-lowering treatment
Glucose	Fasting glucose ≥100 mg/dL, diabetes, or on diabetes treatment
Smoking	Past or current

99.8% had at least one abnormal risk factor in KNHIS and 99.7% had an anbormal risk factor in MESA. For all cardiovascular disease combined (heart attack, stroke, heart failure), the number was 99.3% to 99.5%. More than 93% had two or more abnormal risk factors. Only one person in three or four hundred reached a heart attack with completely clean traditional biomarkers.

Why earlier studies seemed to show the opposite

Older studies often reported that 50% of heart attack patients had “normal cholesterol.” Two methodology choices created this illusion.

Medication masking. If someone has been on a statin for ten years, their LDL at the time of their heart attack might read 90 mg/dL. That looks normal. Their untreated LDL might have been 200. The 2025 JACC study handles this by counting “on lipid-lowering treatment” as an abnormal risk factor in itself. The same logic applies to blood pressure (antihypertensives) and glucose (diabetes medication).

Threshold definitions. As the lead author put it, many risk factors “have a continuous rather than a yes-or-no relationship with risk of developing CVD.” The 2025 JACC study used 200 mg/dL for total cholesterol instead of the older 240 mg/dL cutoff, and 120/80 for blood pressure instead of 140/90.

These two effects reinforce each other. A patient diagnosed with hypertension at age 45, started on a blood pressure medication, and then having a heart attack at 65 with a “normal” reading of 128/82 would be coded as having “no risk factors” in many older studies. In reality, they had two decades of warning.

How the six core heart biomarkers fit in

The 2025 JACC study used four traditional risk factors. Three of those map closely to three of the six biomarkers we cover in the best biomarkers for heart disease in 2026:

JACC traditional factor	More accurate biomarker
Blood pressure	Blood pressure
Total cholesterol	ApoB
Fasting glucose	HbA1c

ApoB directly counts atherogenic particles, which is why it outperforms LDL for heart attack prediction. HbA1c reflects average blood sugar over two to three months, instead of a single fasting snapshot. Substituting these in only sharpens the 99% finding.

The remaining three biomarkers (Lp(a), hs-CRP, and eGFR) are not part of the 2025 JACC paper. Each catches a different mechanism:

• Lp(a) is the strongest hereditary risk factor, set mostly at birth.
• hs-CRP measures the inflammatory pathway, which is independent of cholesterol.
• eGFR captures kidney function, a force multiplier for cardiovascular risk.

What “predictable” means for your own risk

99% of heart attacks have warning signs. The question is whether you’ll see them in time.