Microbiomes Beyond the Gut: Skin, Mouth, and the Soil Under Our Boots
Written By Kevin Wallace.
The gut gets most of the attention, and it should. It is the largest microbial community we carry, and it is the one a kitchen can change. It is not the only one. Skin, mouth, and the ground we work are separate habitats. They talk to the immune system. They do not follow the same rules, and a product labeled “microbiome” does not make them one system.
This is the chapter after the gut piece. Same place, Oswego County, zone 5. Same limit: what a person can actually do this week, and what the studies do not yet show.
1. Where the skeptics are right
A lot of the marketing is ahead of the data. “Balancing your microbiome” is not a diagnosis. Swab tests sold to consumers still do not tell you which lotion to buy or which supplement will fix eczema. Skin communities differ by site on the same person — oily face, moist armpit, dry forearm — more than a single product can address (3).
Soil contact is not a treatment. The better human trials are short, mostly in children, and they measure immune markers and skin bacteria, not fewer cases of asthma or diabetes (12) (13). Eating dirt is not the intervention. Raw manure and unwashed produce still carry real pathogens. Tetanus is still a reason to keep the vaccination current if you work ground.
Antibacterial everything is not careful hygiene. The FDA’s 2016 final rule pulled 19 ingredients, including triclosan and triclocarban, from consumer wash products because the makers had not shown those soaps were safe for long daily use or better than plain soap and water at preventing illness (14). Handwashing stays. The fragrance-and-triclosan claim does not.
So the skeptics are right about the products. They are wrong if the next sentence is that only the gut matters.
2. Three habitats, not one
The NIH Human Microbiome Project sampled healthy adults at oral sites, skin, nose, stool, and, in women, the vagina. Communities were specialized. Oral and stool samples were diverse in membership. Skin was more variable person to person. What was stable was the job list at a given site, not a shared species roster you could buy in a bottle (1) (2).
| Habitat | What shapes it | What the evidence supports | What it does not support |
|---|---|---|---|
| Mouth | Sugar frequency, saliva, plaque, nitrate-rich greens, smoking, dental care | Less frequent free sugar; greens and a less processed pattern can lower gum inflammation in small trials (6) (8) | A probiotic mint replacing a cleaning or a filling |
| Skin | Site, pH, sebum, washing, antibiotics, occlusion | Plain soap, fewer unnecessary antibiotics, barrier left intact (4) (14) | A gut-strain cream as a cure for eczema |
| Soil contact | Diversity of vegetation and soil, time outside, early-life exposure | Weeks of microbially richer play surfaces shifted skin bacteria and immune markers in children (12) (13) | Ingesting soil, or expecting compost to reseed an adult gut |
3. Mouth: sugar frequency, and greens that are not a garnish
The mouth is the microbiome a dentist can see. Calculus on old teeth held a record of it. Adler and colleagues sequenced ancient dental plaque from 34 European skeletons and reported that the shift from hunting and gathering to farming moved the oral community toward a more disease-associated pattern, and that cariogenic bacteria became dominant around the Industrial Revolution. Modern plaque was less diverse than the historic samples (5). That is a record of diet change, not a prescription to eat like a forager. It does line up with what cavities still track: frequent free sugar, and not enough time between exposures for saliva to recover.
CDC’s 2024 oral-health surveillance report, from NHANES 2017 through March 2020, puts untreated decay in permanent teeth at about one in five adults aged 20 to 64, and about one in ten adolescents aged 12 to 19 (10). WHO’s fact sheet still treats oral disease as a major noncommunicable problem worldwide, driven in large part by sugar, tobacco, and uneven access to care (11). Diet is one lever. It is not a substitute for a cleaning.
Two small randomized trials from Freiburg are the cleanest diet experiments I know of on gums. In 2016, ten people in the experimental arm ate a pattern lower in processed carbohydrate and higher in omega-3 fat, vitamin C, vitamin D, antioxidants, and fiber for four weeks. Plaque scores stayed about the same. Gingival inflammation and bleeding on probing fell to roughly half of baseline, and that drop differed from controls (6). A 2019 trial in thirty gingivitis patients, who were told to stop interdental cleaning, found the same clinical pattern: bleeding down, plaque not the explanation. Subgingival community composition did not change in those four weeks (7). Read that carefully. The gums calmed. The swab did not rewrite itself.
Nitrate is a more specific oral story. A trial in periodontal recall patients gave a lettuce-juice drink with about 200 mg of nitrate a day for two weeks after cleaning. Gingival inflammation fell, and nitrate-reducing genera, Rothia and Neisseria, rose in the test arm, not in the placebo arm (8). A 2024 systematic review of nitrate and the oral microbiome found the same direction across studies: health-associated Neisseria and Rothia up, periodontal-associated Prevotella down in a minority of studies (9). Leafy greens and beets are the food form. A juice trial is not a reason to swallow nitrate salts.
Here that is ordinary food, not a protocol. Kale, chard, and lettuce if the season allows, cooked greens in winter, beets in the root cellar. The limit is the same as the gut article: the plant, not a powder with a strain name on it.
4. Skin: leave the barrier alone more often than you treat it
Skin is a dry, acidic, aerated surface. The organisms that live there are not the colonic fermenters. Reviews from the Segre and Gallo groups describe a community that occupies follicles and the stratum corneum, trains local immune cells, and occupies space a pathogen would otherwise take (3) (4). Atopic dermatitis flares are associated with Staphylococcus aureus overgrowth. That association is real. A cream of gut lactobacilli is not the demonstrated fix.
I used this comparison on the channel talking about brown rot on a peach: intact skin plus a resident community is why ordinary contact does not become an infection, and blasting a surface with a broad killer does not tell you what grows back (16). The human version is plainer than a foliar spray. Wash with plain soap when the hands are dirty. Skip consumer antibacterial washes. Do not put antibiotics on unbroken skin because a label said “antibacterial.” Moisturizer on cracked hands in a zone 5 winter is barrier care, not a microbiome product. If a rash is spreading, that is a clinic visit, not a ferment.
5. Soil: the organisms in the bed are for the bed
On the food-forest side of this property the goal is fungal-dominant ground: wood chips, leaf mulch, chop-and-drop, less bare gravel. The driveway stays bacterially dominated because we drive on it. I have walked through that succession on the channel (15). Those organisms mineralize residue, build aggregates, and hold water. They are not a human probiotic.
The human signal from soil is exposure, mostly early, mostly through skin and airways. In a Finnish daycare trial, children whose yards were rebuilt with forest floor, sod, and planters showed shifts in skin bacteria and a higher plasma IL-10 to IL-17A ratio over 28 days (12). A later double-blind trial used playground sand enriched with diverse soil against visually similar, microbially poor sand. After 28 days the intervention skin community still differed from baseline, and regulatory T cells and cytokine patterns moved with the exposure, not with the look of the yard (13). That supports the biodiversity hypothesis as a hypothesis with a human experiment behind it. It does not say an adult who mulches a bed will reset an immune disease.
The useful version here is ordinary contact while planting and turning compost, then washing up, not a ritual and not a mouthful of bed soil.
6. Honest caveats
Ancient plaque studies are small and European. Diet changed with farming and industry, and so did hygiene, crowding, and flour texture. Adler shows a shift. It does not isolate one food (5).
The gum-diet trials are pilots. Fifteen people in the 2016 experimental comparison, thirty in 2019. Four weeks. The 2019 microbiome result was null even though bleeding fell (6) (7).
Nitrate drinks were given with professional cleaning, in recall patients, for two weeks (8). Greens are food. They are not periodontal therapy.
Soil trials are in young children, with immune-marker endpoints. No trial here shows that garden contact prevents autoimmune disease in adults (12) (13).
Skin probiotic cosmetics are mostly marketing. Strain, dose, and site are rarely the ones used in the clinical papers (4).
We eat a whole-food plant-based pattern here. That is a disclosure, not a finding that every other pattern fails the mouth or the skin.
7. Where we land
The gut chapter still stands: feed colonic microbes with a variety of fibers from real plants. Skin, mouth, and soil need different work. None of it is a cleanse.
Keep free sugar infrequent. A sweet with a meal is a different exposure than sipping it all afternoon. Saliva needs the gap.
Put nitrate-rich greens on the plate several days a week. Kale, chard, lettuce, beets. Cooked is fine in January. This is the oral food with an actual trial behind it (8).
Do not drop the brush, the floss, or the cleaning. In the diet trials, plaque did not fall. Inflammation did, and only partly (6).
Wash dirty hands with plain soap and water. Skip consumer antibacterial soaps (14).
Treat cracked winter skin as a barrier problem. Petrolatum or a plain ointment beats a “probiotic” cream with no strain data.
Get soil on the hands in the garden, then wash. Do not eat the compost. Do not put manure tea on food crops close to harvest.
Build the bed for the plants. Wood chips and fungal compost belong in the food forest. They are not a human supplement (15).
If something in the mouth is bleeding for weeks, or a skin infection is spreading, that is a dentist or a physician, not a longer article.
8. Further Reading
The papers worth sitting with are the Human Microbiome Project overview, the Adler calculus letter, the two Woelber trials, and the Finnish yard studies. They are shorter than the marketing, and they say less.
9. On Our Website
The gut piece is the parent chapter. The soil pieces are the property side of the same boundary: organisms in the bed are there to cycle nutrients, not to be swallowed.
Are Carbohydrates Essential? The Gut Microbiome Says Yes — fiber, short-chain fatty acids, and what we actually eat here.
What Your Gut Microbes Actually Need — the September 30 article.
Johnson-Su Bioreactor: Fungal Compost Without the Turning — how we make fungal compost, and why it stays in the soil program.
Soil Succession: From Barren Soil to a Thriving Forest Ecosystem — bacterial gravel to fungal beds, without the human-health overclaim.
