Can Heat-Killed Probiotics in Toothpaste Reduce Gingivitis?

A bacterium does not need to be alive to affect your gums. In a small crossover trial, adults using toothpaste containing heat-killed Lacticaseibacillus paracasei GMNL-143 had improvements in gingival inflammation and bleeding markers versus placebo treatment periods. The finding is unusual because a heat-killed bacterium cannot colonize the mouth or keep producing metabolites. Its proposed job is much simpler: bind oral pathogens, interfere with their attachment, and possibly engage the mucosal immune system.

That result fits a broader, still uneven clinical literature. Randomized trials of live probiotic lozenges, an oral supplement, and an oral rinse have also reported lower gingival index scores or less bleeding. The shared direction is encouraging. It does not show that every probiotic works, that delivery format does not matter, or that probiotics can replace brushing, flossing, and professional dental care. The studies use different strains, populations, endpoints, and follow-up periods, so they cannot answer those questions cleanly.

What the heat-killed toothpaste trial actually found

Lee et al. (2024) ran a randomized, double-blind, placebo-controlled crossover trial in 30 adults. Participants used heat-killed GMNL-143 toothpaste and placebo toothpaste in separate treatment periods. The study measured gingival index, plaque index, bleeding on probing, and salivary secretory IgA. It reported improvement in clinical gingivitis markers with the paraprobiotic toothpaste.

The most interesting evidence sits beside, rather than inside, the clinical endpoint data. In laboratory experiments, the researchers removed surface-layer proteins from the heat-killed cells with guanidine hydrochloride. Intact cells coaggregated with five oral pathogens—Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, Streptococcus mutans, Fusobacterium nucleatum, and Prevotella intermedia—more effectively than surface-protein-depleted cells. That supports surface-layer proteins as a plausible physical capture mechanism. Adhesion assays also suggested that the cells could temporarily interfere with pathogen attachment to epithelial cells.

Those experiments establish biological plausibility; they do not prove that coaggregation caused the clinical changes in participants. The trial also measured six pre-specified periopathogens in subgingival plaque and gingival crevicular fluid using species-targeted real-time PCR. This is useful pathogen-focused microbiology, but it is not whole-community 16S rRNA sequencing. It cannot show whether the full oral microbial community reorganized, or whether diversity changed.

The design has a major weakness. Its washout between crossover periods lasted one week, which may have allowed carryover into the second period. The GMNL-143 strain also came from its commercial developer, GenMont Biotech, and investigators had institutional affiliation with that organization. That potential conflict deserves weight when interpreting a 30-person study.

Five live-probiotic trials point in the same direction

The result is not isolated to heat-killed toothpaste. Across several small randomized trials, live strains delivered through oral products also improved at least one gingivitis measure. The common signal is lower inflammation or bleeding, not a demonstrated universal effect size.

In a double-blind trial of 62 adults, Babina et al. (2024) tested Streptococcus salivarius M18 lozenges for 12 weeks, followed by a four-week washout. The study measured gingival bleeding index, gingival index, and plaque accumulation at one, two, and three months and after washout. It reported lower gingival bleeding and plaque with supplementation, using Hedges' g to characterize effects over time. This is the longest intervention and only meaningful post-treatment observation among these core trials, although the available report does not provide the post-washout numerical values needed to judge persistence precisely.

Levi et al. (2023) enrolled 60 people with plaque-induced generalized gingivitis and more than 30% bleeding-on-probing sites at baseline. Participants received Bifidobacterium animalis subsp. lactis HN019 lozenges containing 10⁹ CFU twice daily for eight weeks. Along with clinical indices, the investigators measured inflammatory molecules in gingival crevicular fluid. IL-1α, IL-1β, IL-8, MCP-1, and MIP-1β declined after supplementation. These biomarkers fit an anti-inflammatory mechanism, but they remain biomarkers rather than proof that a specific immune pathway produced the clinical effect.

A 42-participant, triple-blind trial by Modiri et al. (2023) tested an oral rinse containing Lactobacillus plantarum MK06 at 10⁸ CFU/mL for four weeks. It measured gingival index, bleeding on probing, plaque index, and simplified oral hygiene index, and reported improvement in gingivitis outcomes. A rinse has brief contact with oral surfaces, which makes transient interference with pathogen adhesion a plausible explanation. It does not demonstrate durable colonization.

Manek et al. (2025) reported the largest numerical change: in 60 adults taking an oral L. reuteri supplement at 10⁸ CFU per day for 30 days, gingival index fell from 1.83 ± 0.18 to 0.97 ± 0.15, about a 47% reduction (p<0.001). Controls changed from 1.80 ± 0.17 to 1.74 ± 0.16 (p=0.08). That difference is striking, but it comes from a single short study published in Bioinformation and needs independent replication before it should set consumer expectations.

The largest trial in this group, Volgenant et al. (2023), enrolled 117 healthy participants in an experimental gingivitis model. Participants stopped oral hygiene for two weeks while researchers evaluated probiotic L. paracasei LPc-G110 and L. plantarum GOS42 lozenges. The study anticipated only a small-to-medium effect on bleeding on the modified gingival margin. Experimental gingivitis offers tight control, but deliberately induced short-term inflammation in healthy volunteers is not the same condition as chronic plaque-induced gingivitis.

Recovery may matter more than prevention

The cleanest challenge to the "probiotics shield your mouth" story comes from Lundtorp Olsen et al. (2023). In a double-blind, placebo-controlled experimental gingivitis trial, probiotics did not prevent oral microbiota dysbiosis during the period when participants stopped oral hygiene. They supported faster restoration toward each participant's pre-dysbiosis microbiota after hygiene resumed.

This distinction changes the practical interpretation. Probiotics may help the mouth recover once plaque control returns; they have not shown that they can offset sustained plaque accumulation. For a heat-killed product, continued use is especially likely to matter because dead cells cannot establish a self-sustaining population. No core trial demonstrates a durable microbiome shift after supplementation stops.

What pooled evidence adds—and cannot add

Systematic reviews reach a cautious version of the same conclusion. Gheisary et al. (2022) and Benavides-Reyes et al. (2025) found statistically significant adjunctive benefits for periodontal clinical measures, including gingival and bleeding outcomes, while reporting high between-study heterogeneity. A 2025 meta-analysis of lactic-acid-bacteria toothpaste provides format-specific context for toothpaste products.

Heterogeneity is not a footnote. Gingival index, gingival bleeding index, bleeding on probing, and bleeding on the modified gingival margin are related measures, but they are not interchangeable. Strains differ. Doses differ. A lozenge stays in the mouth longer than a rinse, and a toothpaste is used alongside mechanical plaque removal. No trial here directly compares matched strains in toothpaste, lozenges, and rinses within the same population. The apparent cross-format consistency is directional and inferential, not a head-to-head result.

Should you use one?

For an adult with plaque-induced gingivitis, a probiotic or paraprobiotic oral product is a reasonable adjunct if you want to try one and the product identifies its strain and dose. The best-supported choices in this group are S. salivarius M18 and B. lactis HN019 lozenges, which have double-blind trials with longer treatment or immune-marker data. Heat-killed GMNL-143 toothpaste has a clever, plausible mechanism and early clinical support, but its evidence rests heavily on one small crossover trial.

Use probiotics as an add-on, not an alternative to plaque control. Persistent bleeding gums warrant a dental examination, professional cleaning when indicated, and consistent mechanical oral hygiene.

The evidence supports modest confidence that selected probiotic approaches can reduce gingivitis markers during use. It does not establish a durable cure, a replacement for standard care, or a winner among delivery formats. A well-powered trial comparing live and heat-killed products against placebo and established active care, with long follow-up and whole-community sequencing, would answer the questions the current studies leave open.