Which Probiotic Strains Have Evidence for Oral Health

Why the Strain Name Matters

Probiotics are named in three parts: genus, species, and strain. In Streptococcus salivarius K12, "Streptococcus" is the genus, "salivarius" is the species, and "K12" is the specific strain. Two strains of the same species can behave very differently: one may produce an antibacterial compound or stick well to the tongue, while its close relative does neither. That's why a study on one strain doesn't tell you much about another, even if the label shares the first two words.

A small but telling example comes from one of the earliest gingivitis trials. Researchers gave 59 people with moderate to severe gingivitis one of two different Lactobacillus reuteri formulations or a placebo for two weeks. Gum inflammation improved in all three groups (everyone also got brushing and flossing instruction), but only one of the two L. reuteri products improved significantly more than placebo.^[1] Same species, same dose, different result.

A naming note: in 2020 many Lactobacillus species were reclassified into new genera, so you may see Limosilactobacillus reuteri, Ligilactobacillus salivarius, Levilactobacillus brevis, or Lacticaseibacillus rhamnosus on newer labels and papers. The bacteria are the same; we use the older names below because that's how most of the studies describe them.

This article is the evidence map: which strains have been tested, for what, and how well. If you want the practical buyer's checklist (reading a label, dose, CFU counts, manufacturing standards), that's in our guide to how to choose an oral probiotic.

Lactobacillus reuteri for Gingivitis and Periodontitis

Strains studied: L. reuteri DSM 17938 and L. reuteri ATCC PTA 5289, usually given together in a lozenge.

L. reuteri is the most studied probiotic for gum health. Besides the 2006 gingivitis trial above, in which one formulation reduced both gingivitis and plaque over two weeks,^[1] the best-known periodontitis trial tested lozenges containing DSM 17938 and ATCC PTA 5289 (100 million CFU of each) in 30 people with chronic periodontitis. All participants received a professional deep cleaning (scaling and root planing). Half then took the probiotic lozenge twice a day for 12 weeks, and half took a placebo.^[2]

Both groups improved, as expected after deep cleaning. The probiotic group had significantly more pocket depth reduction and attachment gain in moderate and deep pockets, and a larger reduction in Porphyromonas gingivalis, a key gum disease bacterium.^[2]

Evidence quality: Randomized, placebo-controlled trials, which is good, but small (30 and 59 people) and short (2 to 12 weeks). The strongest signal is as an add-on to professional treatment for periodontitis, not as a stand-alone fix.

Streptococcus salivarius K12 and M18

S. salivarius is one of the first bacteria to colonize a baby's mouth and is a normal, abundant resident of the tongue. Two strains have been developed as oral probiotics, and they illustrate strain specificity well: same species, studied for different jobs.

K12 for bad breath. K12 produces bacteriocins, natural antibacterial peptides. In a preliminary study of 23 people with halitosis, everyone first used a chlorhexidine rinse for three days to clear existing bacteria, then took lozenges containing either K12 or placebo. One week in, 85% of the K12 group and 30% of the placebo group had substantial drops in volatile sulfur compounds, the gases behind bad breath.^[3] The authors called it preliminary, and it was conducted by researchers affiliated with the company developing the strain. For more on causes of bad breath, see our bad breath guide.

K12 for throat infections. An Italian study gave K12 slow-release tablets to 31 children with recurrent strep throat for 90 days and compared them with 30 untreated children. Strep episodes in the treated group fell from 94 the year before to 3, versus 90 to 84 in controls.^[4] That's a striking number, but the authors themselves list the limitations: no placebo, no blinding, a small sample, and comparison against the previous year rather than a randomized control group.^[4] Treat it as a promising signal that needs a proper randomized trial.

M18 for cavities and plaque. M18 produces bacteriocins that target Streptococcus mutans, the main cavity-causing species, plus enzymes that may reduce plaque buildup and acidity. In a randomized, double-blind, placebo-controlled trial of 100 cavity-prone children over three months, plaque scores were significantly lower in the M18 group, especially in children who started with high plaque. However, S. mutans counts didn't differ between groups overall; only the subgroup that seemed best colonized by M18 showed lower S. mutans.^[5] Notably, the trial measured plaque and bacterial counts, not new cavities directly.

Evidence quality: K12 halitosis data is small and preliminary; K12 throat data is non-randomized; M18 has one reasonably sized randomized trial with modest results. Several studies involved researchers linked to the strain's developer.

Lactobacillus salivarius WB21 and Lactobacillus brevis CD2

L. salivarius WB21 for bad breath. In a 14-day, double-blind, placebo-controlled crossover trial in Japan, people with oral malodor took tablets containing WB21 (2 billion CFU per day) or placebo, then switched. The results were mixed. Breath odor rated by a trained human judge improved in both the probiotic and placebo periods with no difference between them. But measured sulfur compounds and average gum pocket depth fell significantly more during the probiotic period, along with lower levels of Fusobacterium nucleatum, a bacterium linked to odor and gum disease.^[6] It's a small, short trial with a meaningful but partial result.

L. brevis CD2 for gum inflammation. CD2 carries high levels of an enzyme (arginine deiminase) that may dampen nitric oxide production, one of the drivers of inflammation. In a study of 21 people with chronic periodontitis and 8 healthy volunteers, lozenges containing L. brevis were followed by improvements in clinical signs and lower levels of several inflammatory markers in saliva.^[7] The catch is the design: the study had no placebo group, so it can't separate the probiotic's effect from other influences such as extra attention to oral hygiene during a study.

Evidence quality: WB21 has one small randomized crossover trial with mixed outcomes. CD2's periodontal data here is uncontrolled and should be considered exploratory.

Bifidobacterium animalis subsp. lactis and Lactobacillus rhamnosus GG

B. animalis subsp. lactis: two strains, two different questions. Strain HN019 was tested as lozenges in 41 people with chronic periodontitis, all of whom also had scaling and root planing. Those taking HN019 twice a day for 30 days had significantly greater pocket depth reduction and attachment gain at 90 days than the placebo group, along with fewer periodontal pathogens and lower inflammatory markers.^[8] A different strain, DN-173010, was tested in yogurt in 51 people with healthy gums. After four weeks there were no differences between groups. Only when participants then stopped brushing for five days did the probiotic group show less plaque and gum inflammation than the placebo group.^[9] That's an artificial challenge model, so it tells you about resilience under neglect rather than benefit during normal life.

L. rhamnosus GG (LGG) for cavities in children. This is one of the larger trials in the field. In Finland, 594 children aged 1 to 6 in 18 day-care centers drank either milk containing LGG or regular milk with meals, five days a week, for seven months. The LGG group had less tooth decay and lower S. mutans counts, and their risk of caries was significantly reduced, especially among 3 to 4 year olds.^[10] LGG is best known as a gut probiotic, which is a good reminder that this result came from regular, repeated exposure in milk, and is specific to LGG, not to L. rhamnosus strains in general.

Evidence quality: HN019 has one moderate-sized randomized adjunct trial. DN-173010's result is limited to a no-brushing challenge. LGG has a large, randomized, double-blind trial, but it's a single study from 2001 that hasn't been widely replicated in the same form.

The Evidence Map at a Glance

• L. reuteri DSM 17938 + ATCC PTA 5289: periodontitis (as an add-on to deep cleaning) and gingivitis. Small randomized trials; the most consistent gum data of any strain here.^[2]
• S. salivarius K12: bad breath and recurrent throat infections. Preliminary and non-randomized studies.^[3][4]
• S. salivarius M18: plaque in cavity-prone children. One randomized trial, modest effect, no direct cavity count.^[5]
• L. salivarius WB21: bad breath. One small crossover trial, mixed results.^[6]
• L. brevis CD2: gum inflammation in periodontitis. Uncontrolled study.^[7]
• B. animalis subsp. lactis HN019 / DN-173010: periodontitis adjunct (HN019) and plaque resilience (DN-173010). One trial each.^[8][9]
• L. rhamnosus GG: caries risk in young children via milk. One large randomized trial.^[10]

Across the board, the honest picture is "promising but early." Most strains rest on one or two trials, often small, short, or funded by companies developing the strain. None replaces brushing, flossing, fluoride, or professional cleanings, and the strongest gum results come from probiotics used alongside treatment, not instead of it.

If you're comparing products, it helps to look at the strains on each label and see which outcomes they have been studied for.

See What Is Inside DentaBiome

DentaBiome is a daily oral postbiotic built around three strains, L. plantarum, L. salivarius and L. rhamnosus, plus xylitol, cranberry extract and inulin fiber. L. salivarius is one of the species discussed above. The formula is designed to support gum comfort, fresher breath and a balanced oral microbiome as part of a regular routine, and the 60-day money-back guarantee lets you try the Berry Frost chewable risk-free.

See DentaBiome and the 60-day guarantee →

Affiliate link. The official page has the current packages and refund terms.