Biomarker

For decades, total cholesterol reigned as the biomarker for heart disease risk, driving billions in statin prescriptions. Then researchers discovered the HDL/LDL ratio mattered more than the total number, and later that particle size might matter even more than that. This evolutionary understanding reveals how biomarkers aren't fixed truths but provisional insights that shift as our measurement sophistication grows—what seemed like a definitive answer in 1980 became a crude approximation by 2020.

C-reactive protein (CRP) tells you inflammation is happening somewhere in your body, but it won't tell you if it's your joints, your gut, or your arteries—it's like a smoke alarm that can't tell you which room is burning. This limitation captures the fundamental nature of most biomarkers: they're correlative signals, not causal explanations. When your doctor sees an elevated marker, they're not looking at the disease itself but at molecular breadcrumbs that something, somewhere, might be wrong.

Your lab results show everything in the "normal range," but whose normal? These reference ranges are typically derived from the middle 95% of tested populations—which increasingly means populations already dealing with obesity, prediabetes, and chronic inflammation. What's statistically normal in a sick population isn't the same as optimal health. Functional medicine practitioners have started using "functional ranges" that are tighter than lab normals, recognizing that you can be heading toward disease long before you're technically "abnormal."

Hormone replacement trials initially used cardiovascular events as outcomes, but pharmaceutical studies shifted to using bone density as a biomarker proxy for fracture risk—easier to measure, faster to show results. The problem? Some drugs improved bone density but actually increased fracture risk, revealing that the biomarker and the outcome weren't as coupled as assumed. This happened with diabetes drugs that lowered blood sugar (the biomarker) while increasing heart failure risk (the outcome that actually matters), reminding us that improving measurements isn't always the same as improving health.

An HbA1c of 5.6% might be fine for you but alarming for your friend if her usual baseline is 5.0%—yet standard medicine treats both numbers the same. Longitudinal tracking of your own biomarker trends over time often reveals more than a single snapshot compared to population averages. This is why biohackers obsessively track their data: they're building personal reference ranges, looking for deviations from their own normal rather than asking if they fall within the bell curve of strangers.

The observer effect isn't just for quantum physics—knowing your biomarker numbers changes your behavior, which changes the numbers themselves. People who wear continuous glucose monitors often see their average levels drop not because of medication, but because watching real-time spikes after meals naturally modifies food choices. This feedback loop transforms a passive diagnostic tool into an active intervention, blurring the line between measurement and treatment in ways that challenge traditional clinical trial design.