AI Summary
→ WHAT IT COVERS Peter Attia builds a practical framework for evaluating genetic testing across cardiovascular disease, cancer, neurodegeneration, and pharmacogenetics. He distinguishes high-value tests like BRCA panels from low-value consumer SNP products, explains why phenotype measurement often outperforms genotype prediction, and outlines four questions to ask before ordering any genetic test. → KEY INSIGHTS - **Pre-test framework:** Before ordering any genetic test, answer four questions: What specifically are you trying to learn? Is genetics the right tool, or can phenotype be measured directly? Will the result change your behavior or treatment? Are you psychologically prepared for either outcome? This framework prevents the common mistake of generating data without generating clarity or actionable next steps. - **Consumer SNP tests vs. clinical panels:** Direct-to-consumer genotyping products like 23andMe scan only hundreds of thousands of common SNPs and tested only three BRCA variants originally, while thousands of pathogenic BRCA mutations exist. A negative consumer result does not rule out hereditary cancer risk. For meaningful cancer genetic assessment, clinical-grade panel testing from a CLIA-certified laboratory is required, not a consumer product. - **Phenotype beats genotype for cardiovascular risk:** For most cardiovascular and metabolic disease risk factors, including LDL, ApoB, blood pressure, Lp(a), and insulin resistance, directly measuring the biomarker provides more precise and actionable information than inferring risk from DNA. Lp(a) is almost entirely genetically determined, yet Attia still measures it directly because the measurement delivers more clinically useful data than the genotype alone. - **Two-axis test evaluation:** Evaluate any genetic test on two dimensions: effect size (does the variant dramatically shift risk, like BRCA, or nudge it modestly, like MTHFR?) and actionability (does knowing the result change screening, treatment, or planning?). BRCA and Lynch syndrome score high on both axes. MTHFR and COMT score low on both. Pharmacogenetics scores moderate on effect size but high on actionability. - **MTHFR over-interpretation:** Up to 40% of the population carries MTHFR variants, which alter folate metabolism modestly. The extreme prevalence is itself evidence that average clinical effect is small, since natural selection removes variants causing serious harm. Despite this, MTHFR is routinely used in functional medicine to justify supplement protocols for fatigue, brain fog, and anxiety, conditions with no established causal link to these common variants. - **Pharmacogenetics as highest-utility application:** Pharmacogenetic testing offers the clearest clinical return for common conditions. Roughly 10% of people carry CYP2C19 loss-of-function variants rendering clopidogrel (Plavix) inactive, requiring an alternative antiplatelet drug. HLA-B*5801 carriers face life-threatening hypersensitivity to allopurinol, making pre-treatment testing standard of care. When the question is medication response rather than disease prediction, genetic signal translates directly into a specific clinical decision. → NOTABLE MOMENT Attia describes a patient with HDL cholesterol of 100 mg/dL and LDL of 80 mg/dL who believed his lipid panel indicated low cardiovascular risk for years. A rare SCARB1 mutation was causing falsely elevated HDL. A coronary calcium scan revealed extensive arterial disease, illustrating how specific genetic variants can make standard biomarkers actively misleading. 💼 SPONSORS None detected 🏷️ Genetic Testing, Hereditary Cancer Risk, Pharmacogenetics, APOE Alzheimer's Risk, Cardiovascular Biomarkers, Consumer DNA Tests