High cholesterol on a clean diet almost never means a patient failed at eating. It usually means one of two things the standard panel never measured — inflammation, or a gene that was decided before she was born.
Call her Teresa. She is a composite, built from the patients I see most weeks in San Antonio, not one woman in my chart. Fifty-two. Walks four miles most mornings. Cooks at home. Her total cholesterol came back at 238. Her doctor’s office called it borderline-high and asked her to “watch her diet.” She had already been watching it for six years.
“I don’t eat the things that cause this,” she told me. “So why do I have it?”
That question is more common than most doctors let on. The CDC’s most recent national data put it at roughly ten percent of American adults twenty and older walking around with total cholesterol at or above 240, the clinical cutoff for “high,” and a far larger group, something like 86 million adults, sitting above 200. Diet explains part of that. It does not explain all of it.
Here is what a standard lipid panel actually tells you. Total cholesterol. LDL, the particle that deposits plaque. HDL, the one that clears it. Triglycerides, which move with sugar and alcohol more than fat. Four numbers. All useful. None of them measure the two things that decide whether a number like Teresa’s is a real threat or a red herring.
The first is inflammation. Specifically, high-sensitivity C-reactive protein, or hsCRP, a marker of the low-grade inflammatory activity happening inside the vessel wall itself, independent of the cholesterol number sitting next to it. This isn’t theoretical. The JUPITER trial, published in the New England Journal of Medicine in 2008, enrolled close to 17,800 people who did not have high cholesterol by the standard definition. What they had was elevated hsCRP. Rosuvastatin, given specifically for that inflammatory marker, cut major cardiovascular events by 44 percent and all-cause mortality by 20 percent in that group. The trial’s entire premise was that the LDL number alone was missing people who were actually at risk, and finding people whose LDL was fine but whose vessels were quietly inflamed anyway.
The second is genetic, and almost nobody orders it. Lipoprotein(a), or Lp(a), is a cholesterol-carrying particle controlled almost entirely by a single gene on chromosome six. A 2022 review in Current Opinion in Lipidology puts the number at roughly one in five people worldwide carrying an Lp(a) level above the threshold that raises cardiovascular risk. It barely moves with diet or exercise. The number is set early and stays close to flat for most of a person’s life. You can run four miles every morning for a decade and it will not change, because it was never about the miles.
This is the piece that gets lost in both directions. The wellness world tells Teresa her cholesterol is fine, the test is flawed, and she should stop worrying because she’s doing everything right. That’s not true either. If her hsCRP or her Lp(a) is elevated, “doing everything right” on diet is answering a question nobody asked. The other extreme, the fifteen-minute visit, hands her a statin prescription and sends her home with a number lowered and nothing explained. She still doesn’t know why. Fear is part of it too, a drug she doesn’t understand, for a risk she was never shown.
I ordered three more tests for Teresa. hsCRP. Lp(a). ApoB, which counts the actual number of cholesterol-carrying particles rather than estimating their total cargo the way standard LDL does. Her hsCRP came back elevated at 3.1. Lp(a) was normal. The ApoB, though, was high enough to matter even though LDL alone had looked borderline.
Lp(a) matters most when it’s the one piece missing from a family story that otherwise makes no sense. A parent who had a heart attack at fifty-five. A sibling on a statin since their thirties, thin, active, no diabetes, no smoking history, nothing on paper that should have put them there. That pattern is a reason to ask for Lp(a) specifically, once, because it’s set early and rarely needs rechecking after. Most primary care panels don’t include it by default. Most patients have never heard the name.
She was not imagining her effort. The four miles were real and they counted. They counted for her blood pressure, her resting heart rate, her mood, her sleep. They were simply never going to touch the inflammation driving her risk, because walking and inflammation are not the same lever.
We talked about a statin. She was afraid of it — most patients are, usually from something they read, rarely from something their own doctor explained. I told her what the JUPITER data actually showed: a drug, matched to the specific problem she had, cutting real cardiovascular events in people who looked exactly like her on paper. Not a drug for everyone. A drug for her particular number.
Teresa is on a low-dose statin now. Her hsCRP is 0.9. She still walks four miles most mornings, and now the walking and the medication are doing two different jobs instead of one job she assumed would cover both. She did not need to choose between trusting her body and trusting a test. She needed both read correctly, by someone willing to order the two numbers her first panel left out.
If your cholesterol is high and your diet already looks clean, the next step isn’t a stricter diet. It’s asking for hsCRP, Lp(a), and ApoB: three tests most standard panels skip, and the three that usually explain what the first number couldn’t.
Dr. Shiv Kumar Goel, MD, is the founder of Prime Vitality Care in San Antonio, with over two decades in hospital medicine.
