Standard guidelines were calibrated on a population average that may not describe you. Here is where primary care falls short, and what background-aware prevention looks like.
I spent the majority of my career looking at biomedical data. I’ve done this most recently with medical imaging at Butterfly Network but earlier at Verily and Google , where I led a team working on a portfolio of consumer health products. When you build such systems, you learn one hard lesson early: a system only measures what it was designed to. Anything outside its purview stays invisible. The primary care annual physical is a system too. And for most people, it is designed to look for the most acute things only. That is not a knock on primary care physicians, who are doing difficult work inside tight constraints. It is a statement about what the standard physical addresses and what it quietly leaves out.
Walk into most annual visits and the cardiovascular workup is a basic lipid panel: total cholesterol, LDL-C, HDL-C, triglycerides. Useful, but incomplete. Two of the markers most closely associated with early cardiovascular risk usually are not on it.
The first is ApoB, a direct count of the atherogenic particles that actually drive arterial plaque. LDL-C estimates the cholesterol carried inside those particles, which is a proxy. The proxy can read "normal" while particle count runs high, which means two people with identical LDL-C can carry very different risks. The second is Lp(a), a largely genetic risk factor that roughly one in five people carry at elevated levels and that standard panels almost never check. If it is high, it has been high your whole life, and nobody told you because nobody looked.
This is the gap at the center of what the American Heart Association now calls cardiovascular-kidney-metabolic health, or CKM: the recognition that heart, kidney, and metabolic disease share roots and progress together, often silently, years before a standard panel flags anything. The kidney piece is another set of system biomarkers, in addition to the two advanced lipid biomarkers above. Early kidney strain shows up in markers like eGFR and urine albumin creatinine ratio (uACR), which are not part of a routine primary care or cardiovascular specialist check. Research suggests roughly 90 percent of people with chronic kidney disease do not know they have it. You cannot act on a number nobody measured.
Here is the part that gets missed most often. Clinical risk guidelines are built on population averages. The AHA's PREVENT equations are a real step forward, estimating risk across large and diverse datasets. But an average is a story about a population, not a prediction about you. If your family history, ancestry, or biology sits away from that average, the standard model can underestimate your actual risk while telling you that you look fine.
South Asians are a great example: they tend to develop cardiovascular and metabolic disease earlier, and at lower body weights than the populations most risk tools were originally tuned on. The pattern is documented well enough that it has its own name in the literature. Lp(a) elevation, again largely inherited, also varies meaningfully across ancestral backgrounds. Roughly 121 million U.S. adults belong to higher-risk ethnic groups whose risk profile is not well captured by a single baseline. "Your labs are normal" can be technically true and still incomplete, because normal was defined for someone whose background does not match yours.
This is not an argument that everyone needs to panic. It is an argument that the same number means different things for different people, and a screening model that ignores that will reassure exactly the people who should be looking closer.
None of this is a competence problem. It is a system design problem. A primary care visit runs somewhere around 12 to 18 minutes, and it has to cover acute complaints, chronic disease management, required screenings, and prescription refills. There is no slot in that visit for ordering a non-standard panel, interpreting particle counts against your family history, and translating the result into a plan you can actually follow.
The deeper issue is that the system is built to react to disease that has already arrived, not to map risk that is still years upstream. Reimbursement, visit length, and clinical workflow all point in the same direction: treat what is acute, document what is required, move to the next room. So the default response to an early, asymptomatic, background-specific risk signal is the honest answer a rushed system can give: everything looks normal, see you next year.
Meanwhile the numbers keep getting worse. The AHA estimates that around 80 percent of U.S. adults already carry at least one CKM risk factor, and that roughly nine in ten are at stage 1 or beyond. This is not a fringe population. It is most adults, sitting inside a screening model that was never designed to catch them early. Medications like GLP-1s will help bend the curve in terms of how many progress beyond Stage 2, however, these medications are not intended for those in the early stages of CKM disease.
Suppose you do get the right markers measured: an ApoB, an Lp(a), a kidney panel, an HbA1c. Now what? A printout of out-of-range values does nothing on its own. The value is in the translation. What does each marker mean for you specifically? Which ones are modifiable and which need monitoring? What is the concrete sequence of changes that the evidence actually supports for your profile?
The Diabetes Prevention Program trial is the cleanest illustration of the difference. It showed a 58 percent reduction in progression to diabetes from a structured lifestyle program, not from receiving a lab result. The AHA's Life's Essential 8 gives a framework for the levers that move cardiovascular and metabolic risk. The real work is connecting your specific data to those levers, then staying with you while you act on them. That last part, the staying-with-you, is where most consumer testing stops and where measurable risk reduction begins.
Put the pieces together and a different model emerges. Measure the markers a standard panel skips. Read them against your background and family history instead of a generic average. Stage where you sit on the heart-kidney-metabolic spectrum before symptoms show up. Then translate the whole picture into a small number of specific actions, and provide a human who helps you work through them. That is not a more expensive version of the annual physical. It is a different instrument, pointed at a different question: not "are you sick today," but "where is your risk heading, and what changes the trajectory."
This is the problem we built Porter Health to solve. We measure the markers a standard panel skips, we read them against your background instead of a population average, and we turn the result into a prevention plan with specific action items and human coaching to follow through. We are not replacing your physician. We are filling the gap the 15-minute visit was never built to cover: early, background-aware, heart-kidney-metabolic risk awareness that arrives with a plan attached. "Normal labs" should mean you are genuinely low risk. We exist to make sure that is true for you, not just true on average.
