Direct-to-Consumer Genetic Tests: Know Your DNA or Know Your Limits?
What Your Saliva Doesn't Tell You
A tube of saliva, a return envelope, and a few weeks later: a detailed report on your ancestry, disease risks, nutritional sensitivities, fitness traits, and personality tendencies. The appeal of consumer genetic testing is obvious — it promises access to the most fundamental information about who you are and what you're likely to face. But the gap between what these tests measure and what they can responsibly claim is vast, and that gap is where a great deal of marketing lives.
Consumer tests like 23andMe, AncestryDNA, and their competitors work by genotyping — scanning your genome for hundreds of thousands of known single nucleotide polymorphisms (SNPs), specific positions in the DNA sequence where individuals commonly vary. This is real science. What the results mean for your individual health, however, is a different matter.
Most common diseases — heart disease, type 2 diabetes, most cancers — are polygenic: they result from thousands of genetic variants interacting with each other and with a lifetime of environmental exposures, diet, stress, and chance. A polygenic risk score tells you something statistically about population-level associations, but its ability to predict your individual outcome is genuinely limited. A person with a "low risk" genetic profile can still develop heart disease. A person with a "high risk" profile may never do so. The genetic contribution to most common conditions is real but modest — and dwarfed, for most people, by lifestyle and environment.
"When someone reads that they have a 15% lifetime risk versus a 10% population average, they often conclude their genes are working against them. What they're actually looking at is a statistical association derived from GWAS studies with significant population stratification issues, tested on cohorts that are overwhelmingly European in ancestry. The uncertainty bars on that number are enormous."
The ancestry and ethnicity estimates involve similar complexity. They are calculated by comparing your SNP profile against reference populations — panels of individuals whose ancestry is known. The accuracy of these estimates depends entirely on the size and diversity of those reference panels, which are heavily weighted toward European populations. The further your ancestry is from these reference populations, the less reliable the estimates become. An Indigenous South American or East African individual may receive ancestry results with significant uncertainty that the sleek interface does not prominently communicate.
FactoraHealth Comparison Table
| The Claim | What the Science Says |
|---|---|
| "Tells you your disease risk" | Provides population-level statistical associations; individual predictive value is limited for most common conditions |
| "Reveals your ancestry precisely" | Estimates depend on reference panels heavily skewed toward European ancestry; accuracy varies significantly by population |
| "Personalized nutrition & fitness insights" | Nutrigenomics is a nascent field; most consumer recommendations lack robust clinical validation |
| "Your data is private and secure" | Data breaches, third-party sharing arrangements, and bankruptcy asset sales represent real risks to irreversible biological data |
So What Should We Make of This?
Consumer genetic testing genuinely delivers something: ancestry estimates that many people find meaningful, the occasional medically significant finding (certain BRCA variants, for instance, are tested and reported), and a fascinating window into human genetic diversity. These are real values.
What it does not reliably deliver is the actionable health guidance it is often marketed as providing. The "personalized" nutrition and fitness recommendations rest on an evidence base that does not yet support the confidence with which they're presented. And the data privacy question is not hypothetical — it is a demonstrated, ongoing risk with a population of 15 million people's worth of consequence.
If you're considering a consumer genetic test, the most important questions aren't about the science — they're about the business. Who owns your data? What can they do with it if the company is acquired or goes bankrupt? Can you delete it, verifiably? Your genome will be yours for life. The company holding it may not be.