The global direct-to-consumer (DTC) genetic testing market hit $2.7 billion in 2023, with tens of millions of people having spit into tubes and mailed their DNA to companies like 23andMe and AncestryDNA. Yet questions persist: is genetic testing actually worth it, or is it an expensive way to get information you can't act on? The honest answer depends entirely on what type of test you take, what you're hoping to learn, and what you plan to do with the results.

This article breaks down the real science behind genetic testing — what different test types can and cannot tell you, where the accuracy claims hold up, what the privacy tradeoffs look like, and how to decide whether a DNA test makes sense for your situation.

---

Types of Genetic Tests: Not All DNA Tests Are Created Equal

Understanding what you're buying matters before you spend $99 to $299 on a consumer kit.

Direct-to-consumer (DTC) genotyping — the standard 23andMe or AncestryDNA product — uses a technique called genotyping arrays. These chips scan roughly 600,000 to 700,000 pre-selected positions in your genome (called SNPs, or single nucleotide polymorphisms) out of approximately 3 billion base pairs total. That's about 0.02% of your DNA. These tests excel at ancestry inference and identifying common genetic variants associated with traits and moderate-risk conditions. They are not sequencing your genome.

Clinical diagnostic testing is ordered by physicians when there's a specific medical question — inherited BRCA mutations in a family with breast cancer, for example, or pharmacogenomic testing before prescribing a drug. These tests use more sensitive laboratory methods (often Sanger sequencing or next-generation sequencing of targeted regions), are interpreted by genetic counselors, and are often covered by insurance when medically indicated. The clinical bar for accuracy and reporting is substantially higher.

Pharmacogenomic (PGx) testing analyzes specific genes — primarily CYP2D6, CYP2C19, CYP2C9, and DPYD — that determine how your body metabolizes certain medications. This category has some of the clearest clinical utility of any genetic test: whether you're an "ultrarapid metabolizer" or "poor metabolizer" of clopidogrel (a blood thinner) directly affects which antiplatelet therapy a cardiologist should prescribe.

Whole genome sequencing (WGS) reads all 3 billion base pairs and is increasingly available to consumers through companies like Nebula Genomics and Sequencing.com. At 30x coverage, WGS provides far more data than genotyping arrays, though interpreting all that data remains the primary challenge. Most variants of uncertain significance (VUS) returned from WGS don't have clear clinical meaning yet.

---

What DTC Genetic Tests Can Actually Tell You

Consumer genetic tests provide genuinely useful information in several specific domains.

APOE and Alzheimer's risk. 23andMe's health reports include APOE ε4 status, which is the strongest known genetic risk factor for late-onset Alzheimer's disease. Carrying one copy of ε4 increases risk approximately 3-fold; two copies increase risk roughly 8-12 fold. This isn't a diagnosis or certainty — most APOE4 carriers never develop Alzheimer's — but it's one of the most actionable pieces of genetic information a consumer test can return, particularly for lifestyle decisions around cardiovascular health, which is deeply intertwined with Alzheimer's risk.

BRCA carrier screening. 23andMe tests for three specific BRCA1/BRCA2 variants common in the Ashkenazi Jewish population. If positive, this finding warrants clinical follow-up. However, there are over 1,000 pathogenic BRCA variants; the consumer test detects only three. A negative result on 23andMe is not a clinical negative — it means you don't carry those three variants.

Pharmacogenomics. CYP2C19 variants affect how people process clopidogrel, omeprazole, and certain antidepressants. CYP2D6 affects metabolism of codeine, tamoxifen, and many psychiatric medications. DTC tests cover the most common variants in these genes. For people on polypharmacy regimens or anyone being prescribed drugs with narrow therapeutic windows, knowing your metabolizer status is practically useful.

Nutrigenomics. Variants in MTHFR (folate metabolism), FTO (appetite regulation and obesity risk), LCT (lactase persistence), and CYP1A2 (caffeine metabolism) are detectable by consumer tests. The clinical weight of most nutrigenomics findings is moderate — MTHFR variants, for instance, are extremely common and their relationship to homocysteine levels and health outcomes is more nuanced than the supplement industry suggests — but they can inform dietary choices.

Ancestry. Reference panels have expanded substantially. Ancestry composition reports are reasonably accurate for broad continental ancestry and increasingly useful for regional ancestry within Europe, Africa, and East Asia. These are probabilistic estimates, not definitive genealogical records.

---

Where Genetic Testing Falls Short

The limitations are real and often undersold at the point of purchase.

Polygenic traits remain difficult to predict. Height is determined by thousands of genetic variants, yet the best polygenic scores for height still explain only about 40-50% of heritable variation. Complex diseases — type 2 diabetes, schizophrenia, coronary artery disease — are similarly polygenic, meaning no single variant tells you much about your actual risk. DTC tests reporting "elevated risk" for these conditions based on a handful of SNPs should be interpreted cautiously.

False positives in rare variant reporting. A 2018 study by Tandy-Connor and colleagues examined 49 patients who had received DTC results suggesting pathogenic variants in disease genes. When these results were confirmed in clinical labs, 40% were false positives — the DTC test reported a pathogenic variant that either didn't exist or was incorrectly classified. Genotyping arrays are optimized for common variants and have lower reliability when calling rare variants. This is why clinical validation remains essential for any DTC finding that would influence medical decision-making.

Missing variants. Because DTC genotyping covers only pre-selected positions, it will miss pathogenic variants that aren't on the chip. The BRCA example above illustrates this: a negative consumer test does not rule out BRCA-related hereditary cancer risk.

Ancestry results are estimates. Reference panels are not globally comprehensive. People with complex admixture backgrounds, indigenous ancestry, or origins in underrepresented populations get less precise estimates. The field is improving, but this remains a real limitation.

Environmental and behavioral factors dominate most health outcomes. Genetics explains perhaps 30-50% of the variance in most common chronic diseases. Smoking, physical activity, diet, socioeconomic factors, and access to healthcare frequently matter more than any individual genetic variant. A "low genetic risk" result shouldn't create false reassurance about lifestyle.

---

Accuracy: What the Science Actually Says

The accuracy picture is nuanced by what "accurate" means in this context.

For the variants that DTC genotyping arrays are designed to detect — common SNPs — the concordance rate with clinical testing is high, generally above 99% at the variant level. The platforms use well-validated chips with extensive quality control.

The problem arises at interpretation and rare variant calling. The Tandy-Connor 2018 paper found 40% false positive rates specifically for rare variant calls in disease genes — situations where the array either miscalled a variant or where variant classification was incorrect. This doesn't mean genotyping is generally inaccurate; it means rare pathogenic variants are a specific failure mode.

Clinical whole exome or genome sequencing, performed in CLIA-certified labs, has lower false positive rates for rare variant detection. ACMG (American College of Medical Genetics) guidelines govern which variants are reportable and how they should be classified. The Green et al. 2013 ACMG recommendations established a framework for returning "secondary findings" — incidental discoveries of pathogenic variants in genes like BRCA1, BRCA2, and LDLR even when these weren't the primary test question.

For pharmacogenomics specifically, accuracy for the most common variants (CYP2C19 *2, *3, *17; CYP2D6 *4, *5) is generally high from both DTC and clinical platforms, though rare gene duplications (relevant for CYP2D6 ultrarapid metabolizers) can be missed by array-based methods.

---

The Psychological Impact of Genetic Results

Research on how people respond to genetic test results offers a more complicated picture than either genetic determinism optimists or privacy advocates tend to acknowledge.

Bloss and colleagues published findings in the NEJM in 2011 showing that genome-wide profiling for disease risk in generally healthy adults did not produce significant increases in anxiety, distress, or changes in health behavior at one year follow-up. This was a randomized controlled design and represented genuinely good news for the concern that consumer genetic testing would cause widespread psychological harm.

However, the picture isn't entirely reassuring. Roberts and colleagues (2017) found that motivations for DTC testing varied substantially — curiosity, ancestry, health optimization — and that perceived utility depended heavily on whether results were "actionable" in the user's view. People who received APOE4 positive results showed mixed reactions ranging from constructive lifestyle modification to unproductive anxiety about a risk factor they couldn't change.

"Genetic determinism" — the belief that a genetic variant deterministically causes a disease — remains a common misinterpretation. Variants confer probabilistic risk modifications, not certainties. A positive BRCA carrier result means elevated lifetime risk, not certain cancer. An APOE4 result means elevated Alzheimer's risk, not a diagnosis. Genetic counseling helps contextualize these probabilities, but most DTC customers never access it.

---

Privacy, Data Security, and the 23andMe Problem

Privacy concerns around DTC genetic testing are not hypothetical.

Genetic data is among the most sensitive personal information that exists. It's immutable, identifies biological relatives (not just you), and has implications for insurance, employment, and law enforcement that aren't always well-understood at the point of consent.

GEDMatch, a genealogy database, has been used by law enforcement to identify suspects in violent crimes through familial DNA matching — without the knowledge or consent of the distant relatives whose data was matched. At least 50 cases were solved this way before GEDMatch updated its terms to require opt-in consent.

23andMe's financial difficulties, culminating in its 2025 bankruptcy filing, raised acute concerns about what happens to the genetic data of over 14 million customers when the company undergoes ownership changes. The bankruptcy court sale process involved potential acquisition by buyers with no historical commitment to 23andMe's privacy policies. Several state attorneys general issued advisories recommending customers delete their data.

23andMe and AncestryDNA both offer data deletion, but the mechanics vary. Opting out of research data sharing doesn't necessarily delete all copies of your data. Reading privacy policies carefully before testing remains important.

The Genetic Information Nondiscrimination Act (GINA) prohibits discrimination in health insurance and employment based on genetic information. It does not cover life insurance, disability insurance, or long-term care insurance — gaps that matter if you learn you carry a variant with significant health implications.

---

Cost vs. Value: When Is Genetic Testing Worth the Money?

When it's probably worth it:

• You have a family history of a hereditary cancer syndrome and want to understand your risk before a clinical consultation
• You're about to start a medication with known pharmacogenomic interactions (clopidogrel, warfarin, codeine, certain antidepressants)
• You want ancestry information and the other health data is a bonus
• You're curious about nutrigenomics variants (caffeine, lactose, folate) to inform dietary choices
• You understand the limitations going in

When clinical testing is the right answer:

• A physician has identified a specific genetic question based on your medical history or family history
• You need clinically actionable results (BRCA status before a prophylactic surgery decision, pharmacogenomics before chemotherapy)
• Insurance covers clinical testing in your situation — it often does when there's a medical indication

When to wait:

• You're hoping for a comprehensive health risk assessment that will tell you definitively what diseases you'll develop — no current test can do this reliably
• You have significant health anxiety and limited genetic literacy
• You're making major medical decisions based solely on DTC results without clinical consultation

For most of the genuinely useful pharmacogenomic and nutrigenomics data, raw data from a DTC test — combined with tools like Ask My DNA, which uses AI to help interpret your specific variants in plain language — can provide substantially more value than the basic reports offered by testing companies alone. The data is the same; the interpretation layer matters enormously.

---

Frequently Asked Questions

Are DNA tests accurate enough to trust?

For common variants that major genotyping platforms are designed to detect, accuracy is high — typically above 99% concordance with clinical testing. The important caveat is rare variant calls, where a 2018 study found 40% false positive rates in DTC results. The clinical implication: any DTC result suggesting a rare pathogenic variant in a disease gene should be confirmed with clinical laboratory testing before acting on it. For ancestry and common trait associations, the accuracy is generally reliable within the confidence intervals stated.

Should I get a DNA test if I have a family history of cancer?

If you have a significant family history — multiple first-degree relatives with breast, ovarian, colorectal, or other cancers associated with hereditary syndromes — the right first step is often a clinical consultation with a genetic counselor, not a DTC test. Insurance may cover clinical testing when there's a medical indication, and clinical tests examine more variants than consumer products. That said, a DTC test can be a reasonable first step to assess whether carrier screening is warranted, provided you understand that a negative result doesn't rule out risk.

Can genetic testing tell me what medications to take?

Pharmacogenomic (PGx) testing can identify how you metabolize specific drugs — particularly through CYP2D6, CYP2C19, CYP2C9, and DPYD pathways. This information is directly relevant for medications including clopidogrel, warfarin, codeine, tramadol, and many antidepressants. However, medication decisions involve multiple factors beyond genetics: drug interactions, dosing, your medical history, and your physician's clinical judgment. PGx results inform medication selection; they don't prescribe it.

What happens to my DNA data after testing?

Each company's policies differ, but the general situation is: your raw genetic data is stored by the company and may be shared with research partners if you opt in (often default opt-in). GEDMatch and similar genealogy databases have been used in law enforcement investigations. Deletion requests are possible at 23andMe and AncestryDNA, but the process and completeness of deletion vary. Given 23andMe's 2025 bankruptcy, consumers with existing accounts there should review their current data preferences. Reading the specific privacy policy of any company you're considering, and understanding what you're consenting to, is not optional.

Is whole genome sequencing better than a standard DNA test?

Whole genome sequencing (WGS) reads all of your DNA rather than selected positions, providing far more raw data. The challenge is interpretation: most variants uncovered by WGS are variants of uncertain significance (VUS) — we don't yet know what they mean for health. For ancestry and common trait analysis, standard genotyping provides most of the practically useful information at a lower cost. WGS makes more sense for people with undiagnosed rare conditions, those wanting to store comprehensive genetic information for future interpretation, or research contexts. For the health questions most consumers have, the interpretive gap matters more than the data gap.

---

Educational Content Disclaimer

This article provides educational information about genetic variants and testing methods and is not intended as medical advice. Always consult qualified healthcare providers — including genetic counselors and physicians — for personalized medical guidance. Genetic information should be interpreted alongside your medical history, family history, and professional clinical assessment.

---