The effect of a lozenge combining prebiotic arginine and probiotics on caries increment in children during 10-12 months, a randomized clinical trial

Objectives: To investigate the effect of daily use of a lozenge containing arginine and probiotics for 10 – 12 months on caries increment, gingivitis-and plaque occurrence in children aged 5 – 9 years. Methods: In this placebo-controlled, double-blinded, parallel-grouped randomized clinical trial, 343 children were randomly assigned to one of the study arms (1:1). The intervention group ( n = 172) received a lozenge containing Lacticaseibacillus rhamnosus , LGG ® (DSM33156), Lactobacillus paracasei subsp. paracasei , L. CASEI 431 ® (DSM33451) and prebiotic (arginine 2%). The placebo group ( n = 171) received an identical lozenge without arginine or probiotics. Primary canines and molars, and permanent first molars were examined clinically (d/D = ICDAS 1 – 6 ) and radiographically (d/D = R 1 – 6 ) at baseline and follow-up. Missing (m/M), sealed (s/S), and filled (f/F) surfaces (-s/-S) in both dentitions were also included. Utilizing clinical and radiographic scorings, caries experience was classified as d ICDAS1 – 6 msf-s (primary teeth), D ICDAS1 – 6 MSF-S (permanent teeth), d/ D ICDAS1 – 6 m/M-s/S-f/F-s/S (mixed dentition). A weighted and an unweighted score system was applied. Results: The study was completed by 288 children. The dropout rate was 15%. The increase in Δ mean d IC-DAS3 – 6 msf-s and Δ mean d/D ICDAS3 – 6 m/M-s/S-f/F-s/S was lower in the intervention group ( p = 0.007). No differences were found for gingivitis-and plaque occurrence. No product-related side effects were reported. This study followed ICH-GCP including external monitoring. Conclusion: Daily consumption of a lozenge containing prebiotic arginine and two strains of probiotics showed safe use and statistically significantly reduction in caries incrementbut no effect on the mean plaque or gingivitis occurrence in children. The use of a lozenge with arginine and probiotics combined has a promising potential as a supplementary tool for future management of caries. www.ClinicalTrials.gov (NCT03928587). Clinical significance: The combination of prebiotic arginine and probiotics shows clinical potential as a supplementary approach to toothbrushing with fluoride toothpaste in managing caries increment in children. A new era in the management of caries may be emerging.


Introduction
Although largely preventable, caries remains a major public health issue, affecting 60-90% of school-aged children globally [1].The etiology is explained by a shift in the oral microbiota from a symbiotic balance to dysbiosis [2].Dysbiosis occurs through repeated periods of pH drops (<5.5 for the enamel and <6.5 for the dentin) in the oral biofilm caused by carbohydrate-induced acid production.Eventually, this will lead to the dissolution of the teeth.Regular toothbrushing with fluoride toothpaste is evident as caries prevention, but is not always enough in biofilm control [3,4].Thus, supplementing fluoridated toothpaste with prebiotics and probiotics to combat caries has gained interest during the last decades [5].
Arginine is classified as a prebiotic, which can increase the activity of the arginine deaminase system in biofilm bacteria.Thereby, pH in the biofilm raises, and the composition becomes more similar to individuals with DMFT= 0 [6].Probiotics are live microorganisms.Concerning caries prevention, they contribute by a variety of direct effects on the oral biofilm, depending on the strain, but the precise mechanisms regarding caries prevention are currently not well understood [7,8].
According to meta-analyzes the individual ability of prebiotic arginine and probiotics to prevent and reduce caries has been examined in randomized controlled trials (RCT) but no RCTs combining prebiotic arginine and probiotics have to the authors' awareness yet been conducted [9,10].Prebiotic arginine and probiotics together may be able to reverse the oral biofilm from dysbiosis to symbiosis (Fig. 1).Thus, a combined product may have higher clinical potential and is advocated as the future prevention and management of dental caries [5].
The primary aim of this RCT was to investigate the effect of a lozenge containing two probiotic strains and arginine compared to a placebo lozenge on increment of caries in different severity stages and dentitions in children.The secondary aims were to investigate the effect on caries progression-, stagnation-, and regression along with gingivitis-and occurrence of plaque.Therefore, the null hypothesis for the primary aim states that there is no difference between the groups regarding caries increment.

Study design and recruitment
This RCT was a placebo-controlled, double-blind, parallel-group trial conducted May 2019-October 2021.Children from four Danish municipalities (Fredensborg-Helsingør-Hillerød-Rudersdal) were recruited by three authors (CJP-KRE-AB) when their parents picked them up from school.The four municipalities' drinking water contained between 0.31-0.50ppm fluoride.The intervention period was 10-12 months.The protocol was approved by Region Committees on Health Research Ethics, The Capital, Denmark (H-19,002,678), and registered on www.ClinicalTrials.gov(NCT03928587).The study was performed according to principles of good clinical practice (ICH-GCP), monitored by an external clinical study monitor, and data analyzed by an external statistical firm.

Participants and questionnaires
To be included the children needed to be healthy, aged 5-9 years, able to comply during a dental examination, and willing to consume a lozenge daily.Children with severe medical or physically conditions, or parents with language barriers not able to give informed consent were excluded.At baseline and follow-up, the parents responded to a questionnaire regarding the health, oral habits and dietary of their child along with child-and family related characteristics (Table 2).The response on the baseline health questionnaire determined the children's eligibility.Children meeting the inclusion criteria after obtaining signed electronic informed consent from both parents could participate.Due to the Covid-19 lockdown, 11 children turned ten years old between inclusion and baseline examination.All children received a voucher for The National Aquarium Denmark after the follow-up examination.

Randomization, allocation, and sample size
Each child was randomly allocated to the intervention or placebo group (1:1 ratio).The statistician created the randomization list (SAS proc plan procedure; SAS® V.9.4; SAS Institute Inc; Cary; USA), and a person unaffiliated to the study maintained the key.The randomization was not stratified.A dental assistant or CJP assigned the randomization numbers by the sequence in which the parents completed the baseline questionnaires.Block size was six and not exposed.
Sample size was based on mean dmfs= 1.86 including initial caries lesions on 5-years-old (2018) from two of the participating Danish municipalities [11].With α= 5%, β= 20%, and a dmf-s reduction of at least 1.06 (SD=3.24) in the intervention group, a total of 154 children were needed in each group.With addition of 12.5% dropouts, 350 children were required.
Children and their families, examiners, data collectors, statisticians, staff at the Public Dental Health Services (PDHS), and the sponsor were blinded to the study groups until database lock.The study supply coordinator at the sponsor site was unblinded to perform supply of study products.

Calibration
KRE trained and calibrated CJP in the clinical and radiographical caries scores explained below.After the theoretical calibration, KRE/ CJP each performed two clinical and radiographical examinations on eleven children participating in the study to determine the intra-and inter-reliabilities.

Table 1
Overview of the criteria used for clinical and radiographical registrations of caries-related endpoints along with post hoc analyzes (P1-4).The point systems used for the statistical analyzes are illustrated as W for endpoints using the weighted point system (0-6) and UW for endpoints using the unweighted point system (0-1).
Whitish, requiring airdrying for 5 s. to be identified Whitish, seen on a wet surface Localized enamel breakdown due to caries with no visible dentin Distinct cavity into the dentin extending ≤½ of the surface Extensive distinct cavity into the dentin extending > ½ of the surface Filled surface due to caries on primary or permanent tooth surface m/M Loss of tooth due to caries on primary or permanent tooth surface

NC
New caries registered at follow-up in relation to existing filled surface The following principles were used: a) primary canine, first and second molars were included b) only erupted surfaces of the first permanent molars at baseline were included; c) in case of doubt of the clinical or x-ray score, the lowest score was registered at the examination time; d) in case of presence of both caries and filling on the same surface, the score for caries was registered; e) in case of caries at two sites on the same surface the worst lesion was scored; f) temporary fillings were registered as caries.
At the follow-up examination, additional principles were used for the weighted endpoints: g) filled surface during the study received 4 points unless the baseline point was > 4; thus, the same point as baseline; h) sealed surface received the same point as the baseline in the statistical analyses.*P1: Same as primary endpoint but as unweighted.P2: Same as primary endpoint but without dividing ICDAS 1-2 in brown and white lesions.P3: Same as P2 but unweighted.P4: Same as first secondary but including permanent teeth.**For post hoc two and three, the score 1b/1w was weighted with point 1, while 2b/2w was weighted with point 2.

Clinical examination, compliance, and adverse events
Erupted first permanent molars, and primary canines and -molars were assessed clinically and radiographically using a dental unit at the local PDHS.At baseline and follow-up digital bitewings were taken if available bitewings were more than four months old.Radiolucency was classified in seven severities (Table 1) ranging from no radiolucency (R 0 ) to penetration into the pulp (R 6 ).This was slightly modified from Ekstrand et al. [12]; hence, radiolucency in the enamel-dentin junction (R 2 ), along with radiolucency into the pulp (R 6 ) were added.Plaque and gingivitis were registered on six primary teeth (53/55/64/73/75/84), [13,14].Before caries registration, the teeth were cleaned with a toothbrush.ICDAS was used for clinical caries registration (Table 1).The whole ICDAS scoring system was applied, i.e., ICDAS 1-2 were differentiated based on whether the lesion was white (1w/2w) or brown (1b/2b), referred to in present study as extended ICDAS [15].When a clinical or radiographical lesion was detected, the activity was assessed based on location, color and texture of the clinical lesion, and gingival bleeding [16].Saliva samples were collected at baseline, 2-4 months after baseline, and at the follow-up examination.Saliva analyzes and assessment of lesion activity along with detailed dietary will not be presented here.Trained personnel entered data directly into an electronic-Case-Report-Form (eCRF) at baseline and follow-up examinations.CJP and KRE performed half of the baseline examinations each, due to maternity leave for CJP, and CJP performed all the follow-up examinations.At the follow-up examinations, CJP was blinded to the baseline registrations in the eCRF.
The parents were instructed to give their child one lozenge per day after dinner but before bedtime without chewing.Toothbrushing was recommended twice daily with the toothpaste provided at least 30 min before or after intake of the lozenge.An identical study kit with lozenges, a toothbrush, 1,450 ppm F − toothpaste (Colgate® Maximum Caries Protection), and an appendix with images of probiotic-containing items to avoid throughout the trial was distributed to each child four times.Parents with more than one participant were instructed to additionally mark the lozenge boxes next to the randomization number to avoid mix-up.No further specific oral hygiene or dietary instructions/ restrictions were given.The PDHS provided oral health care regardless of the study activity.Following each examination, a PDHS dentist received a written description of the clinical and radiographical findings for each child in order to provide prophylactical activities or treatments if needed.
A weekly electronic-Patient-reported-Outcomes (ePRO) and phone calls were used to encourage compliance.Unused lozenges were returned to the sponsor in a pre-stamped return envelope to be counted four times during the study period.A monthly ePRO was used to determine any adverse events / side effects and used concomitant medication during the intervention.

Protocol changes
The compliance threshold changed from 75% to 70%, because subjects consuming 5/7 lozenges were labeled as compliant in a weekly compliance report.The ethical committee approved the amendment.

Statistical analyzes
Sample size was calculated using Mann-Whitney-Wilcoxon Tests.Cohen's weighted kappa coefficient was used to measure inter-and intra-reliability for clinical and radiographical registrations for KRE-CJP.Only teeth present at both baseline and follow-up and teeth missed due to caries were included in the analyzes.If a tooth was registered present at follow-up but missing at baseline, the radiographic were used to clarify.
This study focused on mixed dentitions and aimed to investigate the possible effect on (1) different severity stages of caries, (2) changes in caries experience (3) caries progression, stagnation or regression, (4) gingivitis, and (5) plaque.Thus, the study assessed one primary and eight ranked secondary endpoints.Additionally, four post hoc analyzes were established in response to the utilized methodology, explained below.
Most cariological endpoints were based on both clinical and radiographical registrations (Table 1).In case of a lesion was only detected clinically or radiographically the registered score was used.If the severity of the clinical and radiographical scores differed, the highest score was used to characterize each lesion (Supplementary 1).A weighted and an unweighted score system was used to evaluate the outcomes (Table 1).The weighted system aimed to demonstrate the severity of caries development or its sequalae in each child during the intervention.In the weighted index, all clinical and radiographical registrations were transformed into weighted points between 0 and 6 regarding the stage of severity.Four points were given to filled, and three points to sealed surfaces, corresponding to the assumed caries severity.In the unweighted index 0 was given in case of sound surfaces and 1 if there was a lesion or its sequalae regardless the severity (Supplementary 1).The points were summed for each child, both at the baseline and the follow-up examinations.Finally, all endpoints were expressed by dividing the outcome by the individual number of tooth surfaces to ensure comparability between groups, but data was also expressed without dividing.Clinical and radiographical caries registrations combined refers to d ICDAS1-6 msf-s (primary teeth), D ICDAS1-6 MSF-S (permanent teeth), d/D ICDAS1-6 m/M-s/S-f/F-s/S (mixed dentition) and d/D ICDAS1-2w/b,3-6 m/M-s/S-f/F-s/S (extended ICDAS, mixed dentition), joint with its sequalae; missing (m/M), sealed (s/S), and filled (f/F) surfaces (-s/S) on primary and permanent teeth, respectively.For the gingivitis-and plaque indices, scores from all six teeth were summed up and divided by the respective number of index teeth present for each child, both at baseline and follow-up.
The numerical variables were analyzed using an unadjusted nonparametric nominal two-sided Mann-Whitney-U test (p<0.05).The binary secondary endpoints were analyzed using generalized linear models (proc genmod in SAS with dist= bin, link= log).Chi-square or Fisher Exact tests were used for binary/categorical variables.Van Elteren test for nonparametric two-way analyzes was used when adjustments in analyzes were needed.Odds ratio (OR) and Number Needed to Treat (NNT) were also calculated.Data were analyzed for Full Analyze Set (FAS) and Per Protocol Set (PPS) using statistical software (SAS® V.9.4; SAS Institute Inc; Cary; USA.)Data from FAS are reported unless otherwise described.The statistical analyzes plan was finalized before the database lock and unblinding.
Four post hoc analyzes were added.Post hoc 1-3 aimed to evaluate the impact of the weighted score system used along with the extended version of ICDAS, and post hoc 4 aimed to evaluated the secondary endpoint in mixed dentitions: P.1) unweighted Δd/D ICDAS1-2w/b,3-6 m/ M-s/S-f/F-s/S, P.2) weighted, not extended Δd/D ICDAS1-6 m/M-s/S-f/Fs/S, P.3) unweighted, not extended Δd/D ICDAS1-6 m/M-s/S-f/F-s/S, and P.4) unweighted Δd/D ICDAS3-6 m/M-s/S-f/F-s/S.All the analyzes are reported on child level (Table 3).

Clinical trial
Of the 343 included children, 288 (84%) completed the follow-up examination.The number of dropouts were 53 (15%) and 2 were excluded (Fig. 2).At baseline the mean age was 7.2 years, and the mean dmsf-s was 1.23 and the mean DMSF-S was 0.07 (d/D= ICDAS 3-6 /R 3-6 ).The caries experience did not differentiate between the groups at baseline.Table 2 illustrates no statistically significant differences in baseline characteristics between the groups in FAS.However, the use of electric toothbrush was significantly higher in the intervention group but only in PPS (p = 0.02).No significant differences were found in baseline characteristics between the children completing the study and the dropouts (p-values >0.05).The mean lozenge consumption was 80% in the intervention group and 76% in the placebo group.Reasons for non-compliance was not reported systematically but via the compliance phone calls the main reason was due to oversight.In the intervention group 116 (67%) use the provided toothpaste compared to 124 (73%) in the placebo group.
A summary of results for each endpoint and post hoc analyses is listed in Table 3.There was no difference in results when data were expressed with or without dividing with number of tooth surfaces.Thus, only non-divided data are demonstrated in Table 3.The effect of the intervention was evident in the first secondary endpoint, and the second and fourth post hoc analyses.Further, caries related change from baseline (△mean) was higher in the placebo group, except for stagnation and regression endpoints, indicating a positive although nonsignificant tendency of the intervention.
In the primary endpoint (weighted Δd/D ICDAS1-2w/b,3-6 m/M-s/S-f/Fs/S) where all the caries related scores in mixed dentition were used no significant differences were found between the groups (p = 0.22).In the first post hoc (unweighted Δd/D ICDAS1-2w/b,3-6 m/M-s/S-f/F-S), which solely varied from the primary endpoint in its utilization of unweighted score system, did not either yield any statistically differences (p = 0.47).In the first secondary endpoint using unweighted score (unweighted Δd ICDAS3-6 msf-s), the cut of level was set at ICDAS 3 /R 3 , and in primary teeth only, a significant difference was found.As a result, the intervention group exhibited increased △mean d ICDAS3-6 msf-s by 0.09 while the increase was 0.49 in the placebo group (p = 0.007, OR: 0.49 [CI: 0.27;0.88],NNT: 8.6).A significant difference was also found in the fourth post hoc analysis only differentiating from the secondary endpoint by mixed dentition (unweighted Δd/D ICDAS3-6 m/M-s/S-f/F-s/ S) (p = 0.007, OR: 0.49 [CI: 0.27:0.88],NNT: 8.6).The second post hoc analysis (weighted, not extended Δd/D ICDAS1-6 m/M-s/S-f/F-s/S), only differentiated from the primary endpoint by not extending ICDAS 1-2 in brown and white lesions, also showed a significant difference between the groups but only in the PPS (p = 0.046, OR: 0.69 [CI: 0.42:1.134],NNT: 10.8).Adjusted statistical analyses revealed no influence from the significantly higher use of an electric toothbrush that was found in the intervention group in PPS.Regarding the Δ mean gingivitis-and plaque indices (secondary endpoint three and seven), there were no significant difference between the groups.
No serious product-related side effects were reported.There was one probably-related and one possibly-related side effect but both in the placebo group (Fig. 2).

Discussion
The mean dmf-s and DMF-S in Denmark is recognized as one of the lowest levels compared to other countries [17].The national mean dmf-s for the 5-9 years old in Denmark was 1.71 and the mean DMF-S was 0.13 in 2019 [11].In the present study the mean d ICDAS3-6 msf-s was 1.23 and the mean D ICDAS3-6 MSF-S was 0.07 at baseline.The enrolled children are considered as a low caries risk group, since d ICDAS3-6 msf-s and D IC- DAS3-6 MSF-S nearly corresponds to the traditional dmf-s/DMF-S and even lower although the addition of information from bitewings, and the inclusion of sealed surfaces.Despite the low caries risk group, the study showed a significant reduction in caries increment, with intervention as compared to placebo, for the first secondary endpoint and two post hoc analyzes in both primary and mixed dentitions.The mean Δd ICDAS3-6 msf-s (first secondary endpoint) was 0.09 (SD=1.27) in the intervention group versus 0.49 (SD=1.22) in the placebo group corresponding to an 82% reduction in Δmean dmf-s in favor of the intervention group with clinical significance (p = 0.007).The saved 0.40 dmf-s at 10-12 months, corresponds well with findings of mean number of saved dmf-s/year (0.40) and DMF-s/year (0.58) when using 1500 ppm F toothpaste versus none-fluoride toothpaste [18,19].
To detect initial caries lesions both ICDAS 1-2w/b3-6 and bitewings were combined and used during the examinations, as a routine protocol, commonly used in the dental clinics.This should be considered as a strength in the present study as without bitewings, almost half of the lesions may have been unnoticed [20].Despite this only one post hoc analysis in PPS disclosed a significant difference between the groups regarding initial caries lesions.The statistical analyses of the primary endpoint did not show significant results.This could be explained by the power calculation that was based on the traditional mean dmf-s including initial caries lesions from two of the municipalities and not ICDAS.Though initial caries lesions were included, this registration is not mandatory in Denmark, why the study may have been underpowered.For the remaining endpoints they were derived from subgroup analyzes.Another reason could be, that a greater discrepancy is found between examiners in scoring of initial enamel lesions [21], why the use of extended ICDAS 1-2w/b3-6 may have resulted in an even lower level of agreement since the second post hoc (weighted, not extended d/D IC- DAS1-6 m/M-s/S-f/F-s/S) disclosed significant difference (Table 3).However, this might be declined because both inter-and intra-weighted kappa values were excellent.
The lack of significant effect of the lozenge on initial caries lesions is more likely to be discovered due to the intervention length.Radiographical examination in children aged 6-8 years showed that 31% approximal lesions in primary teeth, limited to <1/2 enamel, progressed after 1-year, and reached 60% after 48 months.For permanent first molars it was only 2% after 1-year and 59% after 48 months [22].Thus, the more pronounced effect on ICDAS 3-6 /R 3-6 compared to ICDAS 1-6 /R 1-6 could be due to slower progression in non-cavitated lesions especially with daily use of fluoride toothpaste [23,24].The solubility of the tooth tissue at various pH levels is one reason why dentin lesions (ICDAS 3-6 /R 3-6 ) progress faster compared to enamel lesions (ICDAS 1-2 /R 1-2 ), [25,26].It should be noted that arginine has the effect of raising the pH level in biofilm [27].It is therefore postulated that a prolonged intervention period with the lozenge could improve the efficacy also on initial caries lesions.The intervention length is though

Table 2
Comparison of the individual baseline characteristics between the intervention and placebo groups in FAS (n = 343).For two predictable variables analyzing only first permanent molar teeth n = 313.

C.J. Pørksen et al.
comparable to many other probiotic and arginine studies that have been carried out [5,27].Although the duration time of this study should be long enough for caries to develop or progress, it may be too short in a group with low caries risk.However, the intervention period seemed reasonable since an additional daily routine had to be implemented in the participating families.The intervention period may be considered as a limitation of the study.Future studies combining arginine and probiotics with longer intervention times, would be relevant.
The expectation was that there would be a reduction in plaque indices in the intervention group concomitant with a decrease in caries.However, this hypothesis was not supported by the findings (Table 3).This could be explained by the oral microbiota shifting from dysbiosis towards symbiosis rather than having an impact on the plaque index per se (Fig. 1).This hypothesis needs to be validated in future clinical studies analyzing plaque samples and pH level over time.
Other limitations of the present study could be a) the calibration sessions focused only on caries diagnostics, thus a bias in MIH diagnostics along with disagreement between the extension of a filling could have appeared.But any examiners bias should be equally distributed between the groups, b) the examiners had no influence on the participants' dental care at the PDHS during the intervention period and did not obtain any information in this regard.For example, the examiners were unaware of any caries treatment such as if a sealed surface was performed prophylactically or therapeutically, and c) the accurate time for consumption of the lozenge was not obtained.A smaller setting could be of interest to provide details regarding whether a fixed consumption time yields greater benefits.
A weighted-and an unweighted point system were applied.For example, a lesion scored clinically or radiographically as 3 achieved 3 points in the weighted system and 1 in the unweighted system.The weighted system concept was applied to ensure that any changes in each child and the magnitude of caries development or its sequalae were considered in calculating △values during the intervention period.The effect of the applied weighted system had minimal impact, since only the second post hoc analysis showed a significant difference between the groups compared to unweighted scorings and only in PPS analyze (Table 3).In general, post hoc analyzes were performed to ensure that the outcomes obtained from the study were not affected by the implementation of the new method (weighted system).Moreover, to ensure comparison between groups, all endpoints were divided by the quantity of tooth surfaces.However, it was unnecessary as the number of tooth surfaces were almost equal in the two groups.
The registrations were restricted to primary canines and molars, and first permanent molars as children were expected to exfoliate their primary incisors during the study.Further, the incidence of dental caries on primary incisors is low in Danish children [11]; thus, these teeth were disclosed in the dental examinations.Initially, only 5-7-year-olds were included; however, the age range was expanded early in the study period to 5-9-year-olds to account for short recruiting period.
The probiotics were selected based on a preliminary agar test system, conducted at the sponsors' site in 2007 (internal data /unpublished).
LGG® and L. CASEI 431® were shown to competitively exclude oral pathogens, especially S. mutans.The substance was administrated as a lozenge, to provide additional time to obtain local co-aggregation and competitive exclusion of pathogenic/acidogenic/aciduric bacteria in the biofilm [28].Xylitol was added as filler and sweetener to both lozenges in approximately equal amounts, but it was below the recommended daily dose (6-10 g/day) for caries prevention and below a xylitol effect on dental plaque (>5 g/day) [29,30].Moreover, if xylitol had any effect on caries or plaque, it would be comparable in both groups.
To our knowledge, this is the first RCT examining caries increment with prebiotic arginine and probiotics in combination.No prior arginine or probiotic studies have been conducted in mixed dentitions, and elsewhere different strains were administered.Thus, direct comparisons Table 3 Overview of changes (Δ) from baseline in the intervention and placebo groups for each endpoint and post hoc analyzes, presented for the FAS n = 288.a) Only FAS data on children experiencing progression, stagnation, or regression, and only in permanent molar teeth erupted at baseline.b) Results in PPS n = 249.Mann-Whitney test two-sided P-value at 5% significance level.CI: 95%.Odds ratio (OR), and the Number Needed to Treat (NNT).Bolded figures indicate significant differences.
to this study are difficult.Nevertheless, studies have shown that both probiotic supplements and adding arginine e.g., in fluoride toothpaste have additive caries preventive effect [10,27,31].Notably, probiotics appear to be more effective children with high caries risk [5].The more pronounced effect in the present study on ICDAS 3-6 /R 3-6 is similar to that found in Rodríguez study  ).The Rodriguez study compared milk supplemented with probiotic lactobacilli with standard milk for a similar intervention period of 10 months.However, opposite to the present study, Rodriguez's study was conducted among a population with high caries risk [32].Regarding arginine, a statistically significant effect has been found in studies conducted on permanent teeth but only after 2-years in contrast to the present study.The difference could be explained by the mixed dentition or the combined arginine and probiotics [33,34].Overall parameters of importance for efficacy appear to be duration of the intervention, stages of the lesions, caries risk level, and strains specificity.
In general it is difficult to show a superior anti-caries effect in populations with low to moderate caries risk [35].Thus, one of the major advantages of this study is its capacity to show that combined prebiotic arginine and probiotic is effective in a low caries risk group and within an intervention period of 10-12 months.Due to the low risk of the children in the present study, despite the use of both clinical and radiographical examination, the results are applicable to all caries risk groups.Further, considering an NNT of 8.6 in the current study, it is quite possible that the efficacy would be higher, and NNT lower, in a moderate to high caries risk group.Therefore, future studies with a longer intervention period and involving different caries risk groups are of interest.Nevertheless, the present study shows that a combination of prebiotic arginine and probiotics are a promising supplemental tool for caries management and maybe also as caries prevention.The null hypothesis for the primary aim stating no difference between the groups in caries increment is therefore rejected.

Conclusion
In healthy children with low caries risk, daily consumption of a lozenge combining prebiotic arginine and two strains of probiotics, showed statistically significant reduction in caries increment, but no effect in the mean dental plaque or gingivitis.The combined product seems to be a promising future tool for caries management as a supplementary approach to toothbrushing with fluoride toothpaste.As the first study to investigate the combined effect the results may mark the step stones in dawn of a new era in caries management.

Fig. 1 .
Fig. 1.The etiology of dental caries illustrated as a wheel-like process of dysbiosis.Contrarily, regular pre-and probiotic addition promotes symbiosis and maintains balance in the oral microbiota with no progression or development of dental caries.*Due to probiotics, **Arginine Deaminase System.

Fig. 2 .
Fig. 2. Flowchart (Consort 2010) of the randomization, dropout, exclusion, and compliance of the children in the clinical study.