Precision medicine has been a fashionable phrase for over a decade. Most of what was promised in the early years has not arrived. Whole-genome sequencing was supposed to drive prescribing decisions. It mostly does not. Pharmacogenomic panels were supposed to be standard of care. They are not. Personalized cancer therapy has made real strides. Personalized cardiovascular and metabolic medicine has crawled.
The reason is that genetics turned out to be a weaker driver of disease than the field hoped. Your DNA is the script. Your environment, your microbiome, your sleep, your stress, your nutrient status, and your exposome are what get read off that script and how loudly. The variable layer is where the leverage is, and metabolomics is the read that gets us closest to it.
What is real now
A handful of things in this space have moved from research to clinical utility.
Functional metabolomics. Panels that read organic acids, amino acids, oxidative stress markers, and cofactor status are clinically useful today. I use them weekly. They drive interventions that work.
Microbiome sequencing. Stool panels that read the gut microbiome have moved from research to clinical practice. They are not yet at the resolution required to engineer a microbiome from scratch, but they reliably identify the dysbiosis patterns that matter clinically.
Continuous biometric data. A patient with a wearable now has higher-resolution data on their heart rate, sleep, glucose response, and recovery than most hospital admissions produce. The data is patient-owned and free. The bottleneck is interpretation, not collection.
Targeted pharmacogenomics. A small set of drug-gene interactions is well-enough characterized to change prescribing. SSRIs and CYP450 variants is one. Plavix and CYP2C19 is another. Most of the rest of the field is not there yet.
Advanced lipidology. ApoB and Lp(a) outperform LDL-C in cardiovascular risk prediction by enough margin that any practice doing serious cardiometabolic work should be ordering them.
What is hype
Several claims in this space need to come back to earth.
Biological age tests. Methylation-based age scores produce a single number that satisfies the appetite for a verdict. The number is not actionable. Methylation function, by contrast, is. I read function, not score.
Whole-genome screens for adult disease risk. With a few exceptions (BRCA, Lp(a) genetics, a handful of others), most consumer genome screens add little to family history. The value of the genome is increasing but the clinical utility today is narrow.
AI symptom checkers. Useful for triage. Not a substitute for clinical reasoning. They generate good first hypotheses and miss the patterns that come from sitting with a patient and reading the full picture.
Single-molecule supplements that "fix" complex conditions. NAD precursors, fancy peptides, niche compounds. Most have a real but modest effect within a real protocol. None of them outperform fixing sleep, building muscle, and eating whole foods.
What I think a practice looks like in ten years
The work I do today will be a recognizable version of what most primary care looks like in a decade. A small set of physicians doing high-resolution reading of patient biology, building written plans, iterating on data, and treating the system instead of the symptom. The technology will get cheaper. The interpretation skill will remain the bottleneck.
The big shifts I expect:
- Wearable data integrated into the chart by default. The physician sees your sleep, glucose, training load, and recovery without having to ask for it.
- Functional metabolomics and microbiome panels covered by insurance. Once the cost-effectiveness data piles up, the coverage will follow.
- Continuous risk modeling. Instead of an annual physical, your physician monitors a rolling risk profile and intervenes before drift becomes disease.
- Precision pharmacology. Drug choice and dose driven by genetic and metabolic data rather than population averages.
The center of all of it stays the same: a physician who knows you, reads your biology against optimal, and writes a plan. The technology around that center keeps improving. The center itself does not move.
If that is the kind of medicine you want now, the path in is the Precision Call.
