Direct-to-Consumer Oral Microbiome Tests

Mouth · §211

You spit in a tube, send it off, and three weeks later get a report telling you which bacteria are running your mouth. The science underneath is genuine — the bugs in your saliva really do drive cavities and gum disease, and gum disease really is tied to your heart and brain. The kit on top of that science is a different story: thinly validated, optional in any dental guideline, and selling you supplements at the end of the report.

Test · As-needed Evidence Mixed Chapter Mouth

Modest entry, honestly framed. The kit costs you a hundred or two and a saliva sample; the evidence for it changing anything you do is thin, and the report's product recommendations skew toward what the company sells. The one real payoff is informational — many people read their P. gingivalis number and finally connect bleeding gums to heart and brain risk. If that's what gets you back to the dentist, the test earned its money. If you already floss and go twice a year, the cleaning is doing more for you than the report ever will.

Your mouth holds about 700 known bacterial species, give or take a few hundred fungi and viruses. Most of them are friendly or neutral; a handful — Streptococcus mutans on the sugar-eating side, Porphyromonas gingivalis and friends on the gum-disease side — cause most of the trouble when they get the upper hand. A direct-to-consumer test takes the bacteria swimming in your saliva, reads their DNA, and tells you who's there and roughly in what proportions. The "roughly" is doing real work in that sentence.

Two reading methods dominate. Short-read 16S sequencing copies a small chunk of one bacterial gene and matches it against a reference database. It's cheap and well-understood, but it reliably tells you a bug's genus — the family name — and only sometimes the species Regueira-Iglesias 2023. Shotgun metagenomics reads everything and gets to species or strain level, but it costs more. Bristle uses metagenomics; many cheaper options use short-read 16S without saying so on the box. The distinction matters because species in the same genus can behave very differently — one strain of S. mutans can be a cavity machine while another is mostly harmless.

The story the test sells is built on the ecological plaque hypothesis: a healthy mouth is a stable community, and disease shows up when something — sugar frequency, smoking, dry mouth, a weakened immune system — tips the balance toward bugs that eat acid or thrive without oxygen Marsh 1994, Hajishengallis & Chavakis 2021. Knowing who's in the room sounds like it should help you fix the room. In practice, your dentist already does this — by looking at your gums, probing the pockets between teeth, and checking the X-rays for the kind of bone loss that bacteria cause. The saliva report layers a number on top of those signals; whether the number adds anything is the next question.

How sure can we be of what it says?

Two questions hide inside "is the test accurate?". The first is whether it correctly identifies the bacteria in your spit. The answer there is reasonably yes — saliva samples are reproducible, the lab work is CLIA-certified, and the same person sampled on Monday and Friday will get roughly the same community readout Cabras 2016, Lazarevic 2010. The second question is whether the readout predicts whether you have, or will develop, disease. That answer is much fuzzier.

There's also a third question almost no one asks: does acting on the test change anything? Zero randomised trials compare DTC-test-guided care to standard dental care on any clinical endpoint — not cavity rate, not gum disease severity, not inflammation markers, not the cardiovascular numbers the report keeps hinting at. The category is regulated by the FDA as a general wellness product, which means the agency doesn't really regulate it; CLIA covers how clean the lab is, not whether the interpretation on page 14 of your PDF makes sense Hayasaki & Cohen 2024. So you're paying for a measurement that's probably correct about what it's measuring, sitting under an interpretation that nobody independent has audited.

Why anyone cares about the bugs in your mouth

The stakes aren't really about the test — they're about what the test is pointing at. People treat a bleeding gum the way they treat a paper cut: annoying, ignore it, it'll stop. That instinct gets the story exactly backwards. Severe gum disease isn't a local mouth problem; it's a low-grade systemic infection that runs for years, leaks bacteria into the bloodstream every time you chew or brush, and quietly raises your risk of heart attack, stroke, and dementia.

The European Federation of Periodontology and the World Heart Federation looked at every decent study on this and concluded that severe gum disease independently raises the risk of cardiovascular events — to the point where they recommended dentists actually tell their gum-disease patients about it Sanz et al. 2020. The size of the effect is real: people with serious oral disease are between 1.7 and 7.5 times more likely to develop a major systemic disease over the long run, depending which one you look at Hajishengallis & Chavakis 2021. Population studies make the cleaning side of this concrete — in a Taiwanese cohort of over 100,000 people, those who got regular professional cleanings had a 24% lower heart-attack rate and a 13% lower stroke rate over seven years compared to those who never went Chen et al. 2012. A Korean cohort of 247,696 adults replicated the cardiovascular signal and added that brushing one extra time a day was worth another 9% off the event rate Park et al. 2019.

The Alzheimer's thread is more speculative but harder to unsee. A 2019 study found P. gingivalis, the keystone gum-disease bug, in the brains of dead Alzheimer's patients, along with the toxic enzymes it secretes Dominy et al. 2019. The drug company behind that finding ran a Phase 2/3 trial of a gingipain blocker; it didn't pass its main endpoint, the FDA put it on hold, and the gum-disease-causes-Alzheimer's story remains an active scientific argument rather than a settled fact Detke et al. 2021. But every Alzheimer's patient in that trial showed signs of P. gingivalis infection — the association isn't going away.

Here's the leverage you can act on: the version of you with healthy gums avoids inflammation that doesn't show up on any blood test until decades later. The version with chronic gingivitis lives in a slow simmer most days and doesn't know it — subclinical gum disease, the kind that never announces itself. The hygienist who tells you "your gums bled a little here" is naming, for free, the same thing the microbiome report wraps in a colour-coded chart for $120.

What your dentist already does

A competent dental cleaning answers most of what the saliva test claims to. The hygienist measures pocket depth at six sites per tooth — that's the gold standard for spotting gum disease, more direct than any bacterial readout. They check bleeding on probing, which is the live signal that inflammation is happening right now. Bitewing X-rays catch interproximal cavities that no saliva sample will ever see, because the bug that's eating between two teeth is locked into a biofilm that doesn't shed into spit at any useful rate.

The cleaning itself is the intervention that has actual outcome data behind it — the 13–24% lower cardiovascular event rates from the Taiwan and Korea cohorts above attached to going to the cleaning, not to knowing your microbiome score Chen et al. 2012, Park et al. 2019. A standard two-cleanings-a-year cadence typically costs less out of pocket than two annual microbiome tests, and most dental insurance covers preventive cleanings near-completely.

For one specific case the in-clinic version of microbial testing earns its keep: refractory periodontitis that isn't responding to scaling and root planing, where the periodontist wants to pick the right antibiotic. Tests like OralDNA's MyPerioPath, ordered through a dental office, exist for exactly that decision. That's a different product, a different population, and a different question than the at-home version is solving.

What you actually get for your money

The kit arrives in the mail. You don't eat or brush for about half an hour, spit into a tube up to the fill line, screw the cap on, drop the prepaid box in a mailbox. Total active time is maybe 10 minutes including reading the instructions. Two to four weeks later the results land in an app or a PDF.

The reports look serious. Risk scores 1–10 for cavities, gum disease, halitosis, gum recession. Lists of the bacteria found, colour-coded by team — beneficial, opportunistic, pathogenic. Personalised recommendations: specific mouthwashes, specific oral probiotics, specific dietary tweaks, sometimes with a link to buy them from the company's store. That last part is the conflict to keep your eye on — the recommended product is sometimes a generic best-practice item, sometimes the company's own SKU.

Three things the report wants you to believe

"Your microbiome score is a diagnosis." It isn't. A high P. gingivalis reading in your saliva is not gum disease the same way detectable HPV is not cervical cancer — it's a risk signal, not a confirmation. Gum disease is defined by what's happening at the gum line: pocket depth, attachment loss, bone level. Cavities are defined by holes in teeth. A saliva pool averages bacteria across your whole mouth, so a single inflamed pocket or a single rotting interproximal surface can be invisible in the result.

"More of this bug equals more disease." Bacterial counts and disease don't line up cleanly. The classic cavity bug, S. mutans, is found at meaningful levels in plenty of cavity-free mouths, and at low levels in some cavity-prone ones, because strain-level differences and host factors (saliva flow, immune response, tooth anatomy, genetics) determine how the bug behaves in your specific mouth Marsh 1994. The number on the report is real; the rule turning it into a prediction is the wobbly part.

"This probiotic will fix your community." Maybe, in a way that lasts a week or two. Oral probiotics and most targeted mouthwashes have weak and inconsistent evidence for changing the microbiome long-term — the community rebounds toward whatever your sugar intake, smoking status, and immune state were already selecting for Marsh 1994. The interventions that actually move the ecology are upstream: fewer sugary snacks per day, mechanical plaque removal between teeth, not smoking, professional cleaning when there's calculus to remove.

Three ways this goes wrong in practice. The first is score-chasing without a dentist — a worrying number nudges someone toward the recommended mouthwash and probiotic instead of toward the cleaning that would actually catch the cavity forming between their molars. The second is false reassurance — a clean-looking report from someone with a single interproximal lesion or one bad pocket that's not contributing much to the saliva pool. The third is anxiety with no action — a borderline result, an ambiguous bug list, and no clear next move, because the report is consultation-shaped but no real consultant is on the other end Hayasaki & Cohen 2024.

If the report does nothing more than book your next dental appointment, it was worth the spend. If it routes you toward a supplement instead of toward a hygienist, it was net negative.

Where to look next

The boring interventions that actually move oral and systemic risk are worth more attention than the test: regular professional cleanings (twice a year for most people, more often if you already have gum disease), interdental cleaning every day, and reducing the frequency — not just the quantity — of sugary snacks and drinks. For high-risk readers, periodontal screening (probing) at the dentist is the diagnostic that actually counts. The gut microbiome testing market has a near-identical regulatory and evidence story to this one; if you've thought about that, the lessons port directly.

Related in the handbook

— The bacteria this test counts — P. gingivalis and friends — are exactly the bugs that drive gum disease; that's the real thing it measures.
— The bitewing X-rays at your dentist catch the decay between teeth that no saliva sample can see — part of why the cleaning beats the kit.
— Same playbook as IgG food tests: real biology underneath, thin proof the result changes anything, and a supplement upsell on the last page.
— If the report scares you into action, cleaning between your teeth moves the bacteria numbers more than any supplement it recommends.
— Daily mouthwash reshapes the very bacteria this test reads — including the good nitrate-making ones. It can skew the picture as much as fix it.

1. Substance + claimed effects

Direct-to-consumer (DTC) oral microbiome tests are saliva-collection kits sold to consumers — Bristle, Viome's Oral Health Intelligence, Nebula's bolt-on report, with OralDNA's qPCR pathogen panel as the clinician-ordered cousin. The user spits into a tube, mails it to a CLIA-certified lab, and gets back species-level (or, for short-read 16S, genus-level) abundance estimates of the oral bacterial community, occasionally with fungi Bristle white paper 2022. Reports translate the raw composition into per-condition risk scores (caries, periodontitis, halitosis, gingivitis) plus dietary, mouthwash, and probiotic recommendations. Claimed effects fall into four buckets: (1) earlier detection of dysbiosis before clinical symptoms — pitched as precision dentistry; (2) personalised hygiene/diet/product guidance versus generic floss-and-brush; (3) awareness of the oral–systemic-disease link (cardiovascular, Alzheimer's, diabetes, adverse pregnancy outcomes); (4) longitudinal tracking of intervention response. The substance is the testing product itself, not the underlying oral microbiome science — the question this entry adjudicates is whether the DTC layer adds anything actionable on top of a competent dental exam.

2. Evidence by addressing question

mechanism

What's measured: most DTC tests use short-read 16S rRNA amplicon sequencing (Illumina MiSeq, V3–V4 region typical) or shotgun metagenomics (Bristle's stated method). 16S resolves to genus reliably and species only sometimes; full-length 16S and long-read approaches do better but are rarer in DTC products Regueira-Iglesias et al. 2023. The mechanistic premise is the ecological plaque hypothesis: oral health reflects a stable symbiotic biofilm, and disease emerges when environmental stress (sugar frequency, smoking, dry mouth, immune shifts) selects for acidogenic/aciduric species in caries or for gram-negative anaerobes (Porphyromonas gingivalis, Tannerella forsythia, Treponema denticola, Fusobacterium nucleatum) in periodontitis Marsh 1994, Hajishengallis & Chavakis 2021. The keystone-pathogen model says low-abundance organisms (P. gingivalis) can remodel the whole community and prime systemic inflammation via gingipain virulence factors. Identifying these species in saliva is technically feasible; the open question is whether their presence/abundance in an asymptomatic adult predicts anything individually-actionable.

evidence

Analytical validity (does the test detect what it says it detects) is plausibly OK for established platforms — Bristle's self-published white paper claims >99.9% accuracy for organism detection above the limit of detection Bristle white paper 2022, and salivary microbiomes are reproducible within-subject across days, weeks, and years Cabras et al. 2016, Lazarevic et al. 2010. Clinical validity (does the test discriminate disease) is weaker: Bristle's white paper reports AUC 0.88 for periodontal disease (74% sensitivity, 85% specificity) and AUC 0.81 for caries (69% sens / 75% spec) — best-in-class for a salivary test, the company says, but the data are unpublished outside the white paper and have not been independently replicated. Peer-reviewed predictive studies on the salivary microbiome are "scarce and of low methodological quality" — small N, no validation cohorts, no batch-effect control Regueira-Iglesias et al. 2023. Clinical utility (does acting on the test change patient outcomes) is essentially unstudied — no RCTs compare hygiene/dental outcomes for DTC-test-guided care versus standard dental care. A 2024 regulatory analysis concluded that DTC microbiome tests broadly "lack analytical and clinical validity" and operate in an FDA grey zone exempted as wellness products under the 21st Century Cures Act Hayasaki & Cohen 2024.

practice / clinical consensus

No dental society (AAP, ADA, EFP) recommends DTC oral microbiome testing as part of standard caries or periodontitis assessment. The standard of care for periodontitis diagnosis is probing depth + clinical attachment loss + radiographs + bleeding on probing — clinical and radiographic, not microbiological. The EFP/WHF 2020 consensus on periodontitis and cardiovascular disease, the most thorough recent synthesis, recommends periodontal screening and treatment for cardiovascular risk reduction but does not endorse microbial testing as a stratification tool Sanz et al. 2020. Likewise the 2023 EFP/WONCA consensus across CVD, diabetes, and respiratory disease Herrera et al. 2023. Clinician-ordered tests (OralDNA's MyPerioPath) have a longer track record for adjunctive use when periodontitis is refractory to standard therapy and bacterial-targeted antibiotics are being considered — a narrower use case than the DTC product offers.

practicalities

Pricing as of 2025: Bristle's standalone oral health test ~$120 retail (HSA/FSA eligible); Viome's Oral Health Intelligence ~$200–$300 standalone, ~$350+ bundled with gut/cellular tests; OralDNA panels ~$80–$200 ordered through a dental office. Turnaround 2–4 weeks. Sample is a 30-second saliva spit, no clinician required for Bristle/Viome. Reports run 10–30 pages with risk scores 1–10 and named product/probiotic suggestions (sometimes from the testing company's own store — a flagged commercial conflict). Retesting cadence varies: companies often recommend every 3–6 months at full price to track change.

misconceptions

The most consequential reader-side misreading is that a quantitative microbial readout equals a diagnosis. It doesn't: a moderate P. gingivalis abundance in saliva is not periodontitis any more than detectable HPV is cervical cancer — the disease requires clinical findings (probing depth, attachment loss, bone level). Streptococcus mutans CFU thresholds for caries risk are commonly cited (~250,000–1,000,000 cfu/ml) but no universal cutoff predicts caries reliably across individuals because strain-level pathogenicity, host genetics, salivary flow, and tooth anatomy all vary Marsh 1994. The second misconception is "actionable" recommendations: many DTC outputs (use this probiotic mouthwash, add this prebiotic) reflect commercial inventory rather than evidence-based interventions for the user's specific community. The third is the implicit promise that improving the microbiome on paper changes systemic risk — the oral–systemic association is robust at the population level Hajishengallis & Chavakis 2021, but no trial has shown that DTC-test-guided care moves cardiovascular endpoints.

alternatives

Twice-yearly dental cleanings with probing and bitewing radiographs cover the same clinical questions (caries risk, periodontal status) and have observational evidence linking them to systemic outcomes: a Taiwanese nationwide cohort (n>100,000, mean 7-y follow-up) found regular tooth scaling associated with 24% lower MI risk and 13% lower stroke risk versus never-scaled controls Chen et al. 2012; a Korean cohort (n=247,696, 9.5-y follow-up) replicated the cardiovascular signal for regular professional cleaning plus an additional 9% per-extra-brushing-per-day effect Park et al. 2019. Standard interdental cleaning, fluoride, and dietary sugar reduction operate on the upstream environmental driver of dysbiosis Marsh 1994 — the microbiome shift is downstream of behaviours the reader already knows about. For refractory periodontitis specifically, clinician-ordered microbial testing has a defensible niche (antibiotic selection) but is meaningfully different from the asymptomatic-wellness use case DTC products sell into.

failure-modes

Three patterns recur. (1) Score-chasing without a dental exam: an alarming risk score nudges the user toward product purchases (the test company's recommended mouthwash, supplement, probiotic) instead of toward the dentist who would actually diagnose disease. (2) False reassurance: a "good" microbiome report from someone with subclinical periodontitis or interproximal caries that wouldn't show in a saliva sample. Saliva pools microbes across the mouth — site-specific disease (one inflamed pocket, one decaying interproximal surface) can be invisible. (3) Anxiety spirals from a borderline result that has no clear clinical action. The 2024 regulatory analysis explicitly flags "dignitary and economic harms" from DTC microbiome results consumers can't interpret and providers can't fully validate Hayasaki & Cohen 2024.

stakes

Stakes for the substance (DTC testing) are small: skipping the test doesn't worsen oral or systemic outcomes if standard hygiene and dental care are in place. Stakes for the underlying oral health story are real and deserve naming in the article, because the test packages them as part of the pitch. Severe periodontitis is associated with 1.7–7.5× higher risk of various systemic diseases Hajishengallis & Chavakis 2021; the EFP/WHF consensus is explicit that severe periodontitis independently raises cardiovascular event risk and patients should be counselled accordingly Sanz et al. 2020. P. gingivalis DNA and gingipains have been found in Alzheimer's brains, prompting the (failed-as-an-Alzheimer's-drug but mechanistically informative) atuzaginstat trial Dominy et al. 2019, Detke et al. 2021. So: the oral–systemic link is real; the test is a poor tool for acting on it.

payoff

The most defensible payoff is informational, not clinical: the moment a user reads "your P. gingivalis is elevated, here's why your dentist cares about gum bleeding" is the moment many people first connect oral inflammation to heart and brain risk. This is the awareness lever the testing companies legitimately lean on. Whether the payoff justifies the spend versus a conversation with a hygienist is the entry's central trade-off. Behaviour-change studies on personalised health reports show modest, often-transient effects; no oral-microbiome-specific trials exist to quantify the lift.

out-of-scope

Adjacent entries that should exist or do: dental cleanings cadence; flossing/interdental brushes; fluoride toothpaste; xylitol; periodontal screening (probing); the gut microbiome testing parallel (very similar regulatory and validity story for a more crowded category).

3. Credibility range

Optimist case. The oral microbiome is genuinely informative — decades of culture-based and sequencing work confirm that specific dysbiotic signatures associate with caries and periodontitis Marsh 1994, Hajishengallis & Chavakis 2021. Salivary microbiomes are stable within individuals across days to years Cabras et al. 2016, Lazarevic et al. 2010, so a single sample is a reasonable snapshot. The Bristle white paper's claimed AUCs (0.88 for periodontitis, 0.81 for caries) are in the range of clinically useful screens if independently replicated Bristle white paper 2022. The oral–systemic link is robust epidemiologically and mechanistically Sanz et al. 2020, Dominy et al. 2019, and any tool that nudges people toward better hygiene and more frequent dental visits — observationally linked to 13–24% lower cardiovascular event rates Chen et al. 2012, Park et al. 2019 — pays back its $120 if it changes behaviour. Personalised-data effects on motivation are real if modest. For a high-curiosity user already inclined to act on health data, the test is a defensible $120 nudge.

Skeptic case. No published RCT shows that DTC-test-guided care changes any oral or systemic outcome. Clinical validity figures come from the company's white paper, not from peer-reviewed independent replication, and predictive studies on salivary microbiomes more broadly are "scarce and of low methodological quality" Regueira-Iglesias et al. 2023. Short-read 16S — still the default for many DTC products — resolves only to genus, which is too coarse for species-specific risk claims Regueira-Iglesias et al. 2023. The FDA does not regulate these as medical devices; CLIA covers lab process but not clinical interpretation Hayasaki & Cohen 2024. The "personalised recommendations" attached to results are often commercial inventory (the company's own probiotic, mouthwash) without trial evidence in the user's specific microbial state. Standard dental exam + bitewings + periodontal probing covers the same clinical questions for the cost of a copay, with the cleaning itself carrying observational cardiovascular benefit Chen et al. 2012. The headline payoff — awareness of the oral–systemic link — can be delivered by a one-paragraph article or a competent hygienist for free.

Author's call. Skeptical but not dismissive. The underlying microbiome science is real; the DTC product layer on top is a wellness commodity with thin independent validation and no demonstrated clinical-utility advantage over standard dental care. For most readers the test answers a question their dentist already answers more cheaply and more reliably. The case for testing narrows to specific niches: motivated quantified-self users; readers whose dental care is irregular and who'd use the report as an on-ramp; clinician-ordered use in refractory periodontitis where antibiotic selection benefits from microbial profiling. Evidence score lands around 2; controversy is around 3 because clinicians and microbiologists disagree more than they should about how to handle these tests. Action verb is test (the substance literally is a test) with the article framing the decision honestly. The article's job is to (a) name the real oral–systemic story so the awareness payoff lands, (b) keep the reader from overweighting a non-validated report, and (c) point them to the cheaper standard-of-care interventions that move the same needle.

4. Stakeholder + incentive map

Commercial — testing companies. Bristle, Viome, Nebula, OralDNA Labs. Direct revenue from kit sales, retesting cadence, and downstream commerce (recommended supplements, probiotics, mouthwashes, often the company's own SKUs). Investor incentive to position the test as decision-grade.
Commercial — dental clinics offering in-office testing. Differentiation lever ("we use cutting-edge microbial assessment"); upcharge over standard hygiene visits.
Academic — oral microbiology researchers. Genuine interest in moving the field from associations to clinical tools; some funded by testing companies, which warrants flagging on individual papers.
Professional — dental societies (ADA, AAP, EFP). Conservative on adoption without RCT evidence; protective of clinical-exam primacy.
Regulator — FDA. Treats DTC microbiome tests as low-risk wellness products under software exemptions; minimal oversight Hayasaki & Cohen 2024.
Community — biohackers/quantified-self. Strong appetite for personal data even when clinical actionability is weak; supports the retest-quarterly cadence companies prefer.
Skeptic — microbiology/regulatory ethics scholars. Active criticism of validity claims and consumer-harm potential.

5. Population variability

Test performance and value vary by user. Where it's worth the most: adults with persistent bad breath of unclear origin, recurrent gum inflammation despite good hygiene, or refractory periodontitis under specialist care (where microbial profiling can guide adjunct antibiotics — and a clinician should be ordering it, not the user). Where it's worth the least: adults with healthy gums and recent dental visits — the test is unlikely to change care. Smokers, diabetics, and pregnant readers carry the highest baseline periodontitis risk, where the oral–systemic-link information matters most but where the appropriate next step is a periodontist visit, not a saliva kit. Children/adolescents are excluded from most DTC products (Bristle: 16+). Older adults with polypharmacy-driven dry mouth have shifted microbiomes that current reference databases handle poorly. Saliva sample-quality varies with hydration, recent eating, mouthwash use; companies provide pre-collection instructions of varying detail.

6. Knowledge gaps

What hasn't been done: prospective RCTs comparing DTC-test-guided oral care to standard care on any clinical endpoint (caries incidence, attachment loss, halitosis VAS, cardiovascular biomarker). Independent replication of company-claimed sensitivity/specificity figures in peer-reviewed journals. Cost-effectiveness analysis at population scale. What can't easily be studied: long-tail individual variability in which species cause disease in which hosts, given strain-level heterogeneity. What would change the call: a multi-site validation showing DTC-test-guided protocols outperform standard hygienist care on a meaningful endpoint at 12 months; FDA clearance of an oral microbiome test as a diagnostic device. Until then, the wellness framing wins on marketing but loses on evidence.

Scope vs. brief. The brief named four consequences: oral hygiene decisions, awareness of the oral–systemic link, dental costs, and strength of clinical interpretation. All four are covered end-to-end. Oral hygiene decisions and clinical interpretation get the most weight (mechanism + evidence + misconceptions + alternatives); dental costs get a callout in practicalities; awareness of the oral–systemic link is the load-bearing argument of the stakes section.

Action verb call. Considered decide (genuine trade-off, no consensus). Landed on test because the substance literally is a test, and the article frames the decide-or-skip decision honestly within that. If a reviewer prefers decide, the article body would not need to change.

Category call. oral over screening. Screening would have been defensible — this is structurally a screening tool — but the substance is fundamentally about the mouth, the audience reads it that way, and the cross-link surface is dental.

Longevity score. Rated 1 (marginal), not 0. The test itself does nothing for longevity; the score reflects the indirect nudge toward standard dental care that genuinely is associated with lower CV event rates (Chen 2012, Park 2019). A reviewer who reads this as scoring downstream behavior change rather than the substance itself could push it to 0. I kept the 1 because the awareness-of-link payoff is the test's most defensible benefit and warrants surfacing in the rank card via the longevity pitch.

Use of the Bristle white paper. Cited deliberately and labelled in the prose ("their own validation paper"). Needed the specific sensitivity/specificity numbers, and there isn't an independent peer-reviewed equivalent. The voice flags the provenance every time it appears.

Separate-entry candidates flagged for the backlog:

Gut microbiome testing (DTC). Near-identical regulatory and validity story with a much bigger market (Viome, Ombre, Tiny Health, BiomeSight). Should be its own entry; this one only signposts the parallel.
In-clinic periodontal microbial testing. OralDNA's MyPerioPath and equivalents — different product, different population (refractory periodontitis), different decision (antibiotic selection). The DTC story muddies if you fold them in; they deserve their own entry.
Periodontal screening / probing. The diagnostic that actually counts at the dentist; referenced but not its own entry yet.
Dental cleanings cadence. The intervention with the strongest outcome data (Chen 2012, Park 2019) — should be a flagship entry in the oral category.

Future-link candidates when those entries exist: floss / interdental cleaning, fluoride toothpaste, xylitol, dietary sugar frequency, dental cleanings cadence, periodontal screening, gut microbiome testing.

What was deliberately left out. The deeper microbiology argument about whether P. gingivalis causes Alzheimer's vs. tracks it. Touched in stakes; full treatment belongs in either a periodontitis entry or an Alzheimer's-risk-factors entry. Halitosis-specific testing — the most plausibly useful niche for DTC oral microbiome tests (volatile sulfur compound producers) — is folded into mechanism/evidence rather than getting its own section; the entry stays general-purpose.

Pitch tone. The pitches lean candid rather than promotional because the substance scores modestly and the editorial line is "skeptical but not dismissive." A more enthusiastic framing would mis-sell.