Summary

Lab test kits sold straight to consumers do some things well: access when insurance will not pay, privacy for tests such as HIV, and routine screening when you already know what you want. This piece argues they are the wrong tool for clinical nutrition questions. The lab science is not the problem. The gap is in choosing the tests, timing the draw, and reading the results against your medicines, history and goals.

  • A patient on a GLP-1 weight-loss medicine, a woman in perimenopause and a lean, active founder each need a different starting panel. Each panel is adjusted to the person’s medicines and history.
  • Female hormone results depend on the day of the cycle. Standard tests for testosterone and estradiol are often inaccurate at the low levels found in women. The more accurate method is mass spectrometry. Kits almost never use it.
  • Some tests need a physician’s oversight, and ordering platforms route those through a network of ordering physicians.

Summary added 2 October 2026.

Direct-to-consumer lab testing is a crowded market now. LetsGetChecked, Everlywell, Quest Health, Function Health. What was niche five years ago is now mainstream. Walk into any pharmacy and you can buy a thyroid panel off the shelf. Order online before lunch, swab or finger-prick at home, results in a week.

I understand the appeal. The conventional system is genuinely broken: appointments take months, insurance won't pay for the markers that matter, primary care has 15 minutes and a script pad. DTC labs cut through that. They feel like patient empowerment. And in narrow cases, they are.

But here's what eight years of clinical practice has taught me: for the questions clinical nutrition actually has to answer, DTC kits are the wrong tool. Not because the lab science is wrong. These companies often use the same reference labs (Quest, LabCorp) clinicians use. The problem is upstream of the blood draw. It's the selection of the panel itself. And it's what happens to the results afterward.

Here is the honest case for having a clinician choose and order your labs, instead of a self-serve checkout. A disclosure before going further: my Fullscript practitioner account includes a supplement dispensary that would pay me a margin on supplements sold through it. As of 30 September 2026 the store has made no sales.

The case for DTC labs (and where they fit)

I want to be fair before I argue. DTC labs do three things well:

  • Access. If your insurance won't cover an HbA1c and you can't afford a $300 PCP visit just to get it ordered, $39 for a direct test isn't nothing. That's a real win.
  • Privacy. Some markers, like HIV, STI panels, and certain hormonal tests, patients reasonably want to order without involving their primary care chart.
  • Routine annual screening. If you already know what you're looking for (cholesterol panel, vitamin D, basic CBC), and you don't need clinical interpretation in context, DTC is fine. It's the same blood, the same machine.

So I'm not anti-DTC. I'm anti using a screening tool for a clinical problem.

What clinical nutrition labs actually have to do

Here's where it falls apart. The panels a clinician orders for a nutrition question are not “screening.” They're built to answer specific clinical questions about a specific person, on a specific medication, with a specific goal. Three common situations show the difference:

The GLP-1 patient on month four of tirzepatide

She has lost weight. She feels good. But she also feels tired in the afternoons, her hair is thinning, and she has started skipping resistance training because it does not seem to be doing anything. DTC kits would let her buy a generic "hormone panel" or "general wellness" panel. Neither would catch what's actually happening.

A GLP-1 metabolic baseline panel is a sensible starting point for someone in her clinical context. But starting point is the operative phrase. After reviewing her intake, I would customize it: fasting insulin and C-peptide (to see how the medication is reshaping her insulin sensitivity), comprehensive metabolic panel, full thyroid panel including TSH/Free T3/Free T4 (because rapid weight loss can suppress T3 conversion and that’s why she’s tired[2]), ferritin and iron studies (eating less can mean less dietary iron, and observational studies report lower iron stores in people taking these drugs[3]), vitamin D, B12, magnesium, zinc (the micronutrient floor that gets undercut by reduced food volume), and a high-sensitivity CRP. Plus IGF-1, which tracked lean-mass loss in a study of short, severe energy deficit, though that study was not in people taking these drugs[4].

That's the difference. The panel structure exists; the wheel is not reinvented for each person. But the panel is adjusted for her: markers added (IGF-1, thyroid panel beyond just TSH), some dropped if not clinically relevant, draw timing optimized for accuracy. There is no DTC equivalent of that process because no DTC platform sees her medication, dose, history, or symptoms before pricing the panel.

The woman in perimenopause

She wants to understand her hormones because everything feels off. DTC platforms sell panels labeled "Female Hormone Health": usually a 4-5 marker combination of estradiol, progesterone, FSH, and maybe testosterone (typically by immunoassay). The problem: female hormone testing is extraordinarily time-dependent. Drawing FSH on day 18 of a cycle versus day 3 gives wildly different information. Drawing estradiol when she's still cycling versus when she's amenorrheic means completely different clinical interpretations[5]. And critically: testosterone and estradiol measured by standard immunoassay are notoriously inaccurate at the low concentrations relevant for adult women. The LC-MS/MS (mass spec) assay is the clinical gold standard, but it's almost never what DTC kits use[6].

None of the DTC platforms ask the right questions to time the draw correctly. They'll happily run the panel on whatever day she ships them blood. The result: a chart of numbers that's clinically meaningless because the context wasn't captured.

A full female hormone panel can be built specifically for this clinical scenario. One such panel includes 23 tests / 83 biomarkers: sensitive LC-MS/MS estradiol, total + free testosterone (mass spec), SHBG, DHT, estrone, progesterone, DHEA-S, LH/FSH, prolactin, full thyroid (TSH/FT3/FT4/rT3), AM cortisol, vitamin D, iron studies, B12/folate, fasting insulin, HbA1c, lipid panel, CMP. The starter panel is fixed by clinical context; the timing of the draw and any add-ons are individualized after intake. Here's how it compares to a typical DTC "Female Hormone" panel:

Bar chart of markers covered, typical direct-to-consumer female hormone panel versus one full female hormone panel ordered by a clinician: sex hormones 2 versus 8, thyroid axis 1 versus 4, adrenal 0 versus 2, iron and ferritin 0 versus 3, vitamin D, B12 and folate 0 versus 3, metabolic 0 versus 2, lipids 0 versus 0, inflammation 0 versus 0, liver 0 versus 1
Marker coverage comparison. DTC counts based on representative panels from Everlywell, LetsGetChecked, and Quest Health (averaged). Clinician-ordered panel reflects one full female hormone panel (Quest Diagnostics, serum, with sensitive LC-MS/MS for sex steroids).

The founder who is optimizing

He wants performance data. DTC kits would offer him a generic "executive panel." What he actually needs is different than what an older colleague needs. He's already lean, already exercising, already eating well. The marginal value is in finding the things that aren't already optimized, and the things that are silent.

A metabolic and cardiovascular panel is the better starting point for him. It adds ApoB (a better cardiovascular predictor than LDL-C; the 2026 US multisociety guideline says ApoB testing can be useful once LDL-C and non-HDL-C goals are met[7]), Lp(a) (genetic CV risk that conventional panels miss entirely, and that the 2026 US multisociety guideline says should be measured at least once[8]), homocysteine (methylation status), hs-CRP, fasting insulin, GGT (liver fat before ALT rises), and a comprehensive micronutrient panel. After intake I'd drop the "cortisol" and "DHEA" if a DTC kit had suggested them. Those are noise without 4-point salivary sampling, and a single morning serum cortisol is rarely actionable for a high-functioning client.

The starter panel is different for him than for the GLP-1 patient or the perimenopausal patient because the clinical question is different. That selection is the work. It can't be automated by a website's quiz.

What happens to results: the second gap

Even if the panel selection were perfect, DTC kits fail at the next step: interpretation in context.

When your DTC result lands in your inbox, it includes reference ranges (often lab-statistical, not clinically optimal), a one-paragraph blurb, and an upsell for a supplement or a follow-up "consultation." There's no integration with your medications, your goals, your training load, your sleep, your previous trends. Most importantly: there's nobody who knows you reading it.

The blood draw is the cheapest part. The clinical question that decides what to draw, and the interpretation that turns numbers into a plan: that's where the value lives.

The clinician-ordered workflow, in plain terms

Here is what a good clinician-ordered process looks like:

  1. The clinician reviews your intake. Medications, history, what's prompting the visit, what you're trying to learn. A clarifying question or two may come back before anything is ordered.
  2. The clinician builds the panel. Through a practitioner ordering platform, under the clinician’s own license and scope of practice. Customized, not pulled off a shelf.
  3. The ordering platform sends you a secure link. You pay the lab fee and book a draw at a Quest or LabCorp location near you.
  4. Results go to the clinician first, who reviews them before you meet, flags what matters, puts the rest in context, and arrives with a plan rather than a number list.
  5. You and the clinician walk through it together. Not "here's your printout. Google the things that look red." Actual interpretation, with reasoning and next steps.

You could absolutely order generic panels yourself. You'd also be paying for a generic answer to a specific question.

The scope question: can a dietitian even order these?

Yes, in Florida. The standards of practice for licensed dietitian/nutritionists, a rule adopted under Florida Statute Chapter 468, Part X, say that licensees “may order analyses but may not directly collect human specimen samples.” Where a specific test requires physician oversight (testosterone replacement panels, certain hormone work), practitioner ordering platforms such as Fullscript and Rupa Health route those orders through an ordering-physician network. The clinical interpretation stays with the dietitian.

So the legal infrastructure exists. The question is whether you want a clinician thinking about which markers to draw on you, or whether you want to assemble that judgment yourself off a self-serve checkout.

Bottom Line

DTC labs solve an access problem. They don't solve the clinical problem they're often sold as solving. The selection of the panel, the timing of the draw, and the interpretation of the result: those are the work. Automating away the clinician doesn't make the work disappear. It just makes the patient do it. Most patients shouldn't.

If you want labs for a nutrition question

Ask whoever orders them what question each test is meant to answer and who will read the results with you. Sometimes labs are not yet the right next step, and a good clinician will say so.

Correction, 23 September 2026. An earlier version opened with a market size, $3 billion in 2024 and forecast to roughly double by 2028, and charted it out to 2028. I could not find the report it cited, and none of the U.S. estimates I found supports either number. So the figures and the chart are gone. The argument never rested on them.

Correction, 2 October 2026. Four statements in this article were wrong. It said Florida’s dietetics law, Chapter 468, Part X, lets licensed dietitian/nutritionists order laboratory tests. The statute does not mention laboratory tests. The authority is in a rule adopted under it, Florida Administrative Code rule 64B8-44.007. That rule says licensees “may order analyses but may not directly collect human specimen samples.” The state Board of Medicine adopted the rule and can change it without the Legislature, so the Florida point is narrower than I wrote. The iron sentence cited the SURMOUNT-1 trial, which did not report iron. It also said women on these drugs frequently develop iron deficiency. The evidence is observational and mixed. One 2026 review reports lower ferritin in people taking these drugs, and another calls iron deficiency uncommon. The IGF-1 sentence cited the STEP 1 trial, which did not measure IGF-1, and said lean mass was hemorrhaging. It now cites a study that did measure IGF-1, in a different group of people. The evidence for adding iron studies and IGF-1 to this panel is thinner than the article made it sound. The article also said major societies now recommend ApoB over LDL-C as the primary lipid marker. The current United States guideline does not. The 2026 guideline from the American College of Cardiology, the American Heart Association, the National Lipid Association and other groups says ApoB testing can be useful once LDL-C and non-HDL-C goals are met. I found these errors in a review of my own posts. The case for a clinician choosing the panel does not rest on any of them, and it stands. At the same time, a reference with the wrong title, journal and year was corrected, the line on lipoprotein(a) now quotes the current guideline instead of a grade its cited source never gave, a few patient examples were reworded in general terms or removed, a disclosure was added, and an offer to order labs through my practice was removed.

Sources & further reading

  1. Withdrawn 23 September 2026. The market-size report once cited here could not be located, and the figures it supported were removed (see the correction above).
  2. Reinehr T. Obesity and thyroid function. Mol Cell Endocrinol. 2010;316(2):165-71. Rapid weight loss can suppress peripheral T4-to-T3 conversion, lowering free T3 while TSH and free T4 may remain in reference range, a common cause of post-weight-loss fatigue that single-marker TSH testing misses.
  3. Urbina J, Salinas-Ruiz LE, Valenciano C, Clapp B. Micronutrient and Nutritional Deficiencies Associated With GLP-1 Receptor Agonist Therapy: A Narrative Review. Clin Obes. 2026;16(1):e70070. Users had 26 to 30 percent lower ferritin than comparators taking SGLT2 inhibitors; the authors say the data are observational and causality is not established.
  4. Nindl BC, Alemany JA, Kellogg MD, et al. Utility of circulating IGF-I as a biomarker for assessing body composition changes in men during periods of high physical activity superimposed upon energy and sleep restriction. J Appl Physiol. 2007;103(1):340-346. In 35 healthy men over eight days of hard exercise with energy and sleep restriction, changes in free IGF-I tracked changes in fat-free mass.
  5. Stricker R, Eberhart R, Chevailler M-C, et al. Establishment of detailed reference values for luteinizing hormone, follicle stimulating hormone, estradiol, and progesterone during different phases of the menstrual cycle. Clin Chem Lab Med. 2006;44(7):883-887. FSH/estradiol/progesterone reference ranges vary by an order of magnitude across the cycle; cycle-day-blind testing is clinically uninterpretable.
  6. Rosner W, Auchus RJ, Azziz R, et al. Position statement: utility, limitations, and pitfalls in measuring testosterone, an Endocrine Society position statement. J Clin Endocrinol Metab. 2007;92(2):405-413. And Stanczyk FZ, Cho MM, Endres DB, et al. Limitations of direct estradiol and testosterone immunoassay kits. Steroids. 2003;68(14):1173-1178. Both show how unreliable direct immunoassays are at the low hormone levels found in women; the 2003 study found wide differences between kits, especially at postmenopausal levels.
  7. Sniderman AD, Thanassoulis G, Glavinovic T, et al. Apolipoprotein B Particles and Cardiovascular Disease: A Narrative Review. JAMA Cardiol. 2019;4(12):1287-1295. ApoB outperforms LDL-C as a discriminator of cardiovascular risk. The 2024 National Lipid Association consensus finds clinical value in using apoB alongside a standard lipid panel, and the 2026 ACC/AHA multisociety dyslipidemia guideline (Circulation, 13 March 2026) says apoB testing can be useful once LDL-C and non-HDL-C goals are met.
  8. Reyes-Soffer G, Ginsberg HN, Berglund L, et al. Lipoprotein(a): A Genetically Determined, Causal, and Prevalent Risk Factor for Atherosclerotic Cardiovascular Disease, AHA Scientific Statement. Arterioscler Thromb Vasc Biol. 2022;42:e48-e60. And 2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Dyslipidemia. Circulation. 2026;153(17):e1154-e1276. doi:10.1161/CIR.0000000000001423. It says Lp(a) should be measured at least once.
  9. Florida Administrative Code rule 64B8-44.007, Standards of Practice (in force since 26 April 2021), adopted under Florida Statute Chapter 468, Part X, the Dietetics and Nutrition Practice Act. Paragraph (7): licensees “may order analyses but may not directly collect human specimen samples.” The statute itself does not mention laboratory tests.