Coronary plaque: why three CT scan analyses disagreed

A few weeks ago, physiotherapist and nutrition scientist Simon Hill shared heart scans showing a 48% drop in coronary plaque over 16.5 months. This follow up explains how he got those results, whether the number was even real, and why he started a prescription drug anyway. It is one person's process and not medical advice, so your own numbers belong with your doctor.

How the plaque was measured

Hill had two coronary CT angiograms, the first in December 2024 at age 38 and the second in April 2026 at age 39. Both were done at the same imaging center on the same scanner with the same protocol. These scans quantify soft plaque, calcified plaque and total volume in cubic millimeters.

Three software platforms analyzed the same images. HeartFlow showed a 48% drop in total plaque volume. Qangio showed essentially flat plaque, rising from about 10.4 to 11.5 cubic millimeters. Clearly showed an increase of around 20%. Absolute totals also differed, from about 11.5 on Qangio to 60 on HeartFlow and 74 on Clearly, a six fold spread.

All three agreed on what matters most. His plaque burden is small and none showed rapid progression, defined as percent atheroma volume rising by one percentage point or more per year. Qangio uses a trained human to draw the artery walls, while Clearly and HeartFlow are AI driven. A study from the Miami Heart group warned that AI may overestimate small amounts of soft plaque, and his cardiologist advised anchoring to Qangio.

Why he has plaque at all

Hill ate a standard omnivorous diet until 30, with LDL near 125 mg/dL. After switching to a plant based pattern his LDL fell to the 70s. His point is that plaque reflects cumulative exposure, the area under the curve of how high and for how long ApoB stayed elevated. ApoB counts atherogenic particles and is the best single lipid marker, but genetics, blood pressure, insulin resistance, inflammation and Lp(a) also matter. His Lp(a) is 14, but his father had a heart attack at 41 and his grandfather at 49, and up to half of cardiovascular risk may be genetic.

A New England Journal study of 16,000 adults found that more than a third of men in their 30s had plaque somewhere, and 9% had it in coronary arteries. In the 40s, those numbers jumped to two thirds and 31%.

The protocol between scans

• Diet: high fiber, plant based, 1.2 to 1.6 grams of protein per kilogram, with tofu, tempeh, seitan, legumes, nuts, olive oil and avocado. A higher fat plant pattern gave his lowest ApoB.
• Exercise: resistance training three to four times a week, padel four to five times, and a lab measured VO2 max of 51.
• Sleep: in bed before 10 p.m. for eight hours, dim light before bed, a cool dark room and an Eight Sleep Pod.
• Supplements: none aimed at arteries during the window. He took a multivitamin, creatine, 1.5 grams of EPA and DHA and a prebiotic powder.

Which result he trusts and why he medicated

Using Qangio, soft lipid rich plaque fell about 40% while fibrous plaque and calcium rose, which suggests hardening and stabilization. His calcium score of 4 sits above the 75th percentile for his age, which, under the 2026 ACC and AHA dyslipidemia guideline, supports an LDL target under 70 mg/dL.

Given the GLAGOV trial, where plaque regression grew as LDL fell, he chose to aim for an ApoB of 40 mg/dL or lower. He picked a PCSK9 inhibitor over a statin for adherence, since it is an injection every two weeks, and for side effect profile. He started with one drug and then moved to Repatha because it has outcomes data. He also began 2,400 mg a day of aged garlic extract, whose imaging evidence is weaker.

Comparing with other men his age

The REACT study, which used Qangio, reported a median of 42.6 cubic millimeters among men in their 30s who had any coronary plaque, so his 11.5 sits at the low end. Using Clearly against the Miami Heart Study curves, he lands near the 55th percentile for men his age, while HeartFlow's own database places him at the 80th. None of the three reference groups was built for a 40 year old, which shows how platform dependent even percentiles are. He also noted that the arteries were not measured at quite the same size on both scans, so he reran the numbers with artery size fixed, and the answer did not change.

Practical lessons

• Pick one provider and stay with it.
• Leave 2 to 5 years between scans.
• Be skeptical of dramatic short interval changes.
• Combine lifestyle with medication where indicated.

Conclusion

The headline 48% was not the real story. The real story is that his plaque is low, not rapidly progressing, and that diet alone does not reset a lifelong process.