Can Enzyme Lozenges Break Dental Plaque? The Human Evidence

Enzyme lozenges aim to pull apart dental plaque rather than scrub it off or poison the bacteria inside it. The idea has real biochemical logic. The human evidence does not yet show a reliable clinical benefit.

A 2026 randomized controlled trial by Rikvold et al. tested a lozenge containing mutanase, beta-glucanase, and DNase against placebo during 14 days without self-performed oral hygiene. One lower-dose active group had significantly less plaque at day 7. But the trial missed its pre-specified primary endpoint at day 1, the higher-dose group did not perform better, and gingivitis did not improve at day 14. That pattern makes the day-7 result interesting, not practice-changing.

Plaque is a matrix, not a smear of free-floating bacteria

Dental plaque is a biofilm: bacteria embedded in a sticky matrix that helps them attach to enamel and resist removal. Mechanical cleaning remains the practical way to disrupt that structure. Antibacterial products can reduce plaque too, but they act on microbial chemistry rather than selectively dismantling the material that holds the community together.

The three-enzyme approach targets separate parts of that matrix. Mutanase is intended to break down insoluble mutan glucans that help oral streptococci anchor biofilm to teeth. Beta-glucanase targets soluble glucans. DNase targets extracellular DNA, or eDNA, by hydrolyzing its phosphodiester backbone. eDNA is a recognized structural component of early dental biofilm, so attacking it alongside glucans is a sensible attempt to avoid the redundancy that can defeat a one-enzyme product.

That rationale does not prove that a lozenge can deliver enough active enzyme to the biofilm-enamel interface for long enough to matter in a human mouth. It also does not establish a popular mechanistic claim that plaque eDNA resists standard DNase through Z-DNA or G-quadruplex structures. Those specific conformations may be biologically plausible, but the clinical trial does not test or document them. They should not be used as an explanation for product performance without direct evidence.

The first purpose-powered human test produced a narrow signal

Rikvold and colleagues ran a triple-blind, three-arm parallel trial in 80 healthy participants. Participants accumulated new plaque under supervised conditions for 14 days and did not perform their own oral hygiene. They dissolved an enzyme lozenge or placebo three times daily. The two active formulations contained the same enzyme combination, but Active 2 had three times the total enzyme activity of Active 1.

The design matters because it was powered in advance for this specific triple-enzyme lozenge, using a prior seven-day pilot to estimate the needed sample size. Earlier work had also reported plaque reductions with the same enzyme combination: an eight-day orthodontic mouthwash intervention cited by the investigators and a seven-day lozenge pilot. Those studies support the possibility of an effect across delivery formats, but neither replaces a well-controlled confirmatory trial.

The confirmatory result was mixed. The primary endpoint, plaque measured with the Turesky modification of the Quigley-Hein Plaque Index at day 1, did not differ significantly from placebo. Active 1 showed a statistically significant reduction only at the day-7 secondary endpoint. By day 14, neither active formulation retained statistical significance. The available report does not provide the precise effect estimates, confidence intervals, or p-values needed to judge the size and precision of the day-7 difference.

A secondary endpoint can generate a useful lead, especially when it fits a plausible biological story. It cannot erase a missed primary endpoint. The timing could reflect biofilm development: at day 1 there may be too little plaque for the index to detect a meaningful structural change, while by day 14 the mature matrix may have enough overlapping polysaccharides, proteins, and eDNA to resist the lozenge dose. The authors themselves note that plaque grown in vivo has a more complex, more resilient matrix than in vitro biofilm models.

The missing dose-response is the hardest result to explain

Active 2 contained three times the enzyme activity of Active 1, yet it did not outperform Active 1 at any measured time point. That does not prove the lower-dose day-7 result was false. It does remove a major piece of evidence that would make a causal interpretation convincing. If enzyme activity caused the observed plaque reduction, a higher dose would usually be expected to produce at least some consistent advantage.

The lack of a dose-response has several possible explanations: the dose range may have crossed a plateau, the lozenge may not have delivered proportionally more active enzyme to plaque, or the day-7 finding may reflect random variation. The trial cannot distinguish among those explanations. It does show that "more enzyme" did not translate into better plaque scores in this study.

The clinically relevant inflammatory outcome was also null. The Gingival Index of Löe and Silness showed no significant improvement at day 14. Plaque indices are useful surrogate outcomes, but less plaque matters medically only if it produces durable benefits such as less gingival inflammation or lower risk of periodontal disease. Fourteen days is too short to test prevention of periodontitis, yet it was long enough to test the study's gingivitis measure. That measure did not improve.

Why this does not benchmark against standard plaque control

The study compared the enzyme lozenges with placebo, not with chlorhexidine mouthwash or stannous fluoride. Those are established chemical comparators in de novo plaque research. Without an active comparator, the trial cannot tell us whether the day-7 difference, if reproducible, would be large enough to compete with existing options.

Its no-hygiene design creates another limitation. Suspending brushing and interdental cleaning maximizes plaque accumulation and makes a matrix-targeting intervention easier to detect. It is also far from ordinary life. The relevant question is whether an enzyme lozenge adds anything when people brush effectively, clean between teeth, and receive routine professional care. This trial did not test that question.

The trial was triple-blind, which reduces several forms of bias, but Novozymes A/S, the enzyme supplier and product developer, monitored the study. Commercial involvement does not invalidate the data. It does raise the value of an independent replication, particularly when the primary outcome was null and the only positive result appeared at one secondary time point.

Oral biofilm research should not be mistaken for a product endorsement

The oral microbiome responds to many exposures. A 2026 systematic review of clear aligners found that orthodontic treatment can alter oral microbial communities, while clear aligners may be less disruptive than fixed appliances. An umbrella review of e-cigarettes concluded that vaping may contribute to dysbiosis and biofilm accumulation. These findings show that oral biofilms are responsive systems; they do not validate enzyme lozenges.

For periodontal disease, evidence still centers on removing plaque mechanically and treating established disease with professional care. A 2024 network meta-analysis of probiotics in non-surgical periodontitis treatment evaluated probiotics as an adjunct to professional mechanical plaque removal, with probing depth and clinical attachment outcomes. That is a different intervention and population, but it illustrates the standard an enzyme product must eventually meet: disease-relevant outcomes, not plaque scores alone.

There is also a reason to avoid treating the mouth as a microbial battlefield that needs indiscriminate sterilization. In a 2025 crossover study of dietary nitrate and antiseptic mouthwash, oral microbiome responses and nitric-oxide-related vascular effects differed between younger and older adults. The result does not apply directly to enzyme lozenges, but it reinforces a basic point: altering oral ecology can have consequences beyond plaque mass.

Verdict: wait for replication before buying the mechanism

The enzyme combination has a coherent target map, and the day-7 result warrants follow-up. But the central human trial failed its primary endpoint, lacked a dose-response across a threefold activity difference, lost statistical significance by day 14, and did not reduce gingivitis. Those are not cosmetic caveats.

Do not adopt enzyme lozenges as a substitute for brushing, interdental cleaning, or professional dental care. A useful next trial would be independently run, meet a pre-specified primary plaque endpoint, include an active comparator such as chlorhexidine, demonstrate a dose-response, and measure gingivitis or longer-term periodontal outcomes. Until then, the most reliable plaque strategy remains the unglamorous one: remove it mechanically, every day.