Effects of Fluoride and Calcium Phosphate-Based Varnishes on pH, Lactic Acid and Trace Elements in Saliva: A Randomized Clinical Trial.

Background: When dental plaque is not regularly removed, bacteria break down sugars in the diet forming acids as by-products. Lactic acid is the main acid involved in caries. As acids accumulate minerals are lost from the surface layer of the tooth. The imbalance in demineralization/remineralization favours the loss of calcium and phosphate from the teeth. Saliva contains the most important microelements for the remineralization and maturation of dental tissue and plays a crucial role in maintaining the oral environment. Fluoride is the agent par excellence in preventing and detaining cavities. However, remineralization may be hampered by limited levels of calcium and phosphate, and new products have been developed to ensure a constant supply. Two of the most used products are amorphous calcium phosphate stabilized with casein phosphopeptide (CPP-ACP) and tricalcium phosphate modied by fumaric acid (fTCP). Methods: We conducted a controlled, randomized clinical trial of the effect on the saliva of the application of two varnishes − MI Varnish (CPP-ACP with sodium uoride 5%) and Clinpro White Varnish (fTCP with sodium uoride 5%) – applied every three months in children with a high risk of caries, for 12 months. We included 58 children aged 4-12 years, assigned to control (placebo), Clinpro and MI groups. Baseline and three-monthly saliva samples were taken. We assessed changes in pH, lactic acid concentrations and trace elements in saliva. Results: At 12 months, all groups showed a nonsignicant increase in pH levels and a reduction in lactic acid, which was greatest in the placebo group. There was a signicant reduction in 24 Mg, 31 P and 66 Zn levels in the placebo group (p ≤ 0.05), but not in the other elements studied: 23 Na, 27 Al, 39 K, 44 Ca, 52 Cr, 55 Mn, 57 Fe, 59 Co, 63 Cu, 75 As, 111 Cd, 137 Ba, 208 Pb and 19 F. Conclusions: Neither pH, lactic acid concentrations or most salivary trace elements were useful in dening patients at high risk of caries or in monitoring the effect of MI Varnish were shown the expected benets and risks of the study, after which written informed consent was obtained. A questionnaire was administered in parents containing questions on demographic factors, dietary habits, history of disease, uoride therapy, and use of medication and vitamin supplements (Suplementary le 1). The sample size (n = 19 patients/group) was calculated using the evolution data of the triplaque index® and the existing loss index of children treated by our clinic. An alpha risk of 0.05 and a beta risk of 0.20 (power 0.8) in a bilateral contrast was accepted to detect a minimum difference of 0.13 between two groups, assuming that there were 3 groups, and a standard deviation of 0.09. A loss to follow up of 45% was estimated. The monitoring of the patient was 12 months. the Levene test for equality of variance. Pearson’s chi-square test was used to determine between-group differences in sex and a one-way ANOVA test for differences in age. To determine within-group differences in age by sex we used the Mann-Whitney test. To detect between-group differences in DMFS and dmfs values a Kruskal-Wallis test was performed and within the same group between the T0 and T4 a Wilcoxon test was performed. Differences in concentrations of trace elements, pH and lactic acid between baseline and 3, 6, 9 and 12 months were determined by simple variance analysis of repeated measures. When there were differences between the times, two-by-two comparisons were made using the Holm–Sidak test.


Background
According to the 2015 Global Burden of Disease, untreated caries in permanent teeth were the most prevalent condition worldwide, affecting 2.4 billion people, and untreated caries in deciduous teeth were the tenth most prevalent condition, affecting 621 million children worldwide (1).
Tooth decay is a multifactorial, sugar-dependent disease, mediated by cariogenic bio lm (2). When dental plaque is not regularly removed, bacteria metabolize dietary saccharides (3) leading to an imbalance in demineralization/remineralization dynamics that favors the loss of calcium ions and phosphate from the teeth (4).
Lactic acid, the predominant end product of sugar metabolism, is the main acid involved in caries formation. As acids accumulate in the uid phase of the bio lm, the pH falls, and the surface layer of the tooth is partially demineralized. Once sugars are removed from the mouth by salivary dilution and swallowing, bio lm acids can be neutralized by the buffering action of saliva (2). Saliva, the main vehicle in enamel remineralization, contains the most important microelements for the remineralization and maturation of dental tissue (calcium, sodium, magnesium, zinc, uoride) (5), proteins and glycoproteins, and plays a crucial role in maintaining the oral environment, preserving a neutral pH that enables remineralization (2).
Fluoride is the agent par excellence in preventing and detaining cavities (6). Fluoride varnishes were developed to prolong the contact time between uoride and tooth enamel (7). Systematic reviews, meta-analyses and clinical guidelines worldwide recommend uoride varnishes as caries inhibitors and repairers of initial caries lesions in temporary and permanent teeth (8) (9) (10). As remineralization may be hampered by the limited bioavailability of calcium and phosphate, new products have been developed to ensure their constant supplyment (11). Two of the most used products are amorphous calcium phosphate stabilized with casein phosphopeptide (CPP-ACP) and tricalcium phosphate modi ed by fumaric acid (fTCP) (12) , (13).
Functionalized TCP acts as a low-dose calcium and phosphate delivery system (18) and has been shown to remineralize in vitro (16). Clinpro White Varnish (3M ESPE, Saint Paul, MN, USA) is an fTCP varnish that also contains 5% sodium uoride.
Saliva has been widely studied as a possible indicator of susceptibility to caries, including searching for salivary physiology parameters, antioxidant levels, proteins, microelements and trace elements that may indicate the risk of caries (5). However, no controlled, randomized clinical trials have monitored the effectiveness of calcium phosphate varnishes in children at high risk of cavities according to their in uence on pH, lactic acid concentration and salivary trace elements (15) , (19) , (20).
The objective of this study was to carry out a controlled, randomized clinical trial to study the effects of the coating with MI Varnish and Clinpro White Varnish, applied quarterly to children at high risk of cavities, on the evolution of pH, lactic acid concentrations and salivary trace elements for 12 months.
The null hypothesis was that the application of the two varnishes applied quarterly for 12 months does not change either pH, lactic acid concentrations or levels of trace metals in the saliva of children at high risk of caries.

Trial Design
We conducted a controlled randomized trial with a parallel group design between June 2017 and December 2018. The study was double blinded for patients and the statistical analysis. Dental professionals could not be blinded as the commercial presentation and characteristics of the two varnishes were distinguishable.

Participants
Inclusion criteria were: children aged 4-12 years attending the Integrated Child Dentistry Clinic of the University of Murcia for checkups or dental treatment who presented a high or extreme risk of caries according to the CAMBRA protocol (21).
Exclusion criteria were:1) children who had received uoride varnish or other permanent surface treatment containing uoride in the previous 6 months; 2) children tted with orthodontic apparatus 3) children living in an area with uoridated drinking water; 4) children with moderate or severe uorosis or other morphological or anatomical abnormalities of dental development 5) children with systemic diseases causing physical limitations; 6) children with allergy or proven/suspected sensitivity to milk proteins.
Before the study, an informative lea et was provided to all parents/guardians, who were shown the expected bene ts and risks of the study, after which written informed consent was obtained. A questionnaire was administered in parents containing questions on demographic factors, dietary habits, history of disease, uoride therapy, and use of medication and vitamin supplements (Suplementary le 1).
The study was carried out at the University of Murcia Dental Clinic, Hospital General Universitario Morales Meseguer, Murcia.
The sample size (n = 19 patients/group) was calculated using the evolution data of the triplaque index® and the existing loss index of children treated by our clinic. An alpha risk of 0.05 and a beta risk of 0.20 (power 0.8) in a bilateral contrast was accepted to detect a minimum difference of 0.13 between two groups, assuming that there were 3 groups, and a standard deviation of 0.09. A loss to follow up of 45% was estimated. The monitoring of the patient was 12 months.
Patients were randomly assigned using randomization table.
Clinical evaluation and interventions (Interventions, compliance measure, and clinical visual evaluation time points)

Patient selection
The history and examination were carried out by an experienced pediatric dentist, who underwent three training sessions on written and visual instructions, standardization and calibration with participants in vivo. Caries lesions were recorded using a mirror and a WHO probe according to International Caries Detection and Assessment System criteria (ICDAS II) (22).
Patients meeting the inclusion criteria were randomly assigned to one of the three experimental groups. Verbal and written oral hygiene instructions and dietary advice were given to participants and their responsible adults to facilitate and strengthen preventive measures (Supplementary le 2). Each participant received a uoridated toothpaste with 1450 ppm of uoride (Lacer Junior, Lacer SA, Barcelona, Spain) and a manual toothbrush (Lacer Junior, Lacer SA, Barcelona, Spain) which was changed every 3 months. Instructions were given to avoid other sources of uoride during the study period (environmental products, supplements, professional or other dental products). At each check-up we veri ed that study hygiene guidelines and the use of the study toothpaste were being complied with.
There were ve check-ups: baseline and 3, 6, 9 and 12 months. Fluoride and the prevalence of caries were recorded at baseline and 12 months.

Saliva samples
At the beginning of each check-up, 3.5 ml of unstimulated saliva was collected for 5 minutes in a sterile polyethylene tube. Children had not ingested water or food for 1 hour before the examination. Saliva samples were stored at − 20 o C until measurement.
The schedule at which appointments were made was restricted to 15:00 to 18:00 hours for maximum avoidance of circadian uctuations in the variables studied (23). Saliva collection was always performed in the waiting room to avoid the effect of anxiety on the amount and composition of saliva.
Experimental groups, application of varnishes.
After obtaining saliva samples and before application of the varnishes, the teeth were dried with dry compressed air and isolated with cotton rolls and saliva ejector. Then, 0.25 ml of varnish was applied to the surface of the teeth and allowed to dry for 30 seconds. In the placebo group, distilled water was applied with a brush identical to that used to apply the varnishes. Patients were instructed not to rinse their mouths, not to eat or drink for an hour and not to brush until 4-6 hours after application, in accordance with the manufacturers' instructions. The procedure was repeated every three months for one year ( Fig. 1

Trace elements
We analyzed 2 ml of the homogenized sample using mass spectrometry with inductively coupled argon plasma (ICP-MS Agilent 7900; Agilent Technologies Inc.; CA; USA). Ultrapure water (18.2 M Ω) from a water puri cation system (Milli-Q® Reference A+, Merck Millipore) was used to prepare the standard reagents and solutions: 100 µL of the saliva sample was diluted up to 1 mL with a 2% HNO 3 Suprapur solution in ultrapure water. The samples were introduced in a self-sampling spectrometer by the impulse of a peristaltic pump, and were atomized, ionized and the ions generated detected and quanti ed subsequently by mass spectrometry. The isotopes selected for each element studied were: 23 Na, 24 Mg, 27 Al, 31 P, 39

Statistical analysis
Values are expressed as mean ± standard deviation. The Kokmogorov-Smirnov test was used to determine sample normality and the Levene test for equality of variance.
Pearson's chi-square test was used to determine between-group differences in sex and a one-way ANOVA test for differences in age. To determine within-group differences in age by sex we used the Mann-Whitney test. To detect between-group differences in DMFS and dmfs values a Kruskal-Wallis test was performed and within the same group between the T0 and T4 a Wilcoxon test was performed.
Differences in concentrations of trace elements, pH and lactic acid between baseline and 3, 6, 9 and 12 months were determined by simple variance analysis of repeated measures. When there were differences between the times, two-by-two comparisons were made using the Holm-Sidak test.
A paired t-test was used to analyze the within-group evolution of uoride concentrations between baseline and 12 months when there was normality and a Wilcoxon test when there was no normality. One-way ANOVA was used to detect between-group differences in the same period. A value of p < 0.05 was considered signi cant. The analysis was made using the SigmaStat 3.5 statistical software package (Systat Software Inc., Point Richmond, CA, USA).

Results
Of the 80 patients initially reviewed, only 58 met the inclusion criteria ( Fig. 2 [see Supplementary le 4]). After randomization, 18 were assigned to the control group, 19 to the Clinpro group and 21 to the MI group. Of these, 25 children were lost throughout the study. Of the 33 patients who nally completed the study, 12 were in the placebo group, 10 in the Clinpro group and 11 in the MI group.
There were 16 females and 17 males with a mean age of 7.09 ± 2.55 [4-12 years]. The baseline characteristics of sex, age, age by sex and dmfs of the three groups were similar (Table 2). DMFS was signi cantly higher in the Clinpro group (p = 0.039).

Lactic Acid
All groups showed a similar concentration of lactic acid in time 0. Lactic acid concentrations decreased throughout the study and the decrease was higher among the control group than among the two varnices groups. However, the decrease among the former was statistically non-signi cant. (Table 3 [Supplementary le 5]).

Trace Elements
There was a signi cant reduction in 24 Mg concentrations in the placebo group between baseline, 3 and 6 months versus 9 months, and 3 months versus 12 months; 31 P in the placebo group between baseline versus 3, 6, 9 and 12 months; and 66 Zn between baseline versus 9 and 12 months in the placebo group (

Discussion
The null hypothesis was partially proven, since the application of the varnishes did not change either the pH, the concentration of lactic acid or most trace metals studied.
The age of the children included ranged from 4 to 12 years, a population in which caries prevention is common (21). The varnishes used contained 5% sodium uoride, equivalent to 22,600 ppm or 1.19 M of uoride, and calcium phosphate in two chemical forms: CPP-ACP and fTCP. In vitro studies have demonstrated their preventive and remineralizing capacity as they generate supersaturated calcium and phosphate solutions in bio lm and saliva (16) , (24) , (25).
In vulnerable populations, such as infants and children, saliva is the perfect diagnostic medium due to its non-invasive collection, and easy handling and storage (26). A wide range of saliva biomolecules are related to the physiological state and provide useful data on oral and systemic diseases. We selected pH, lactic acid and some trace elements in total unstimulated saliva as biomarkers of the predisposition to caries and the ability of varnishes to modify the risk (2) , (10) , 27 . Reports (27) have found differences in saliva composition depending on whether there is caries or not, but found no difference between stimulated and unstimulated saliva in children, or according to sex or the dental status.
The baseline pH value was between 7.6 and 7.7, although the patients were at high or extreme risk of cavities, and the value did not change signi cantly throughout the study in any group. Although some reports (28) (29) have recorded lower pH values and observed signi cant differences in pH in unstimulated saliva in children without versus children with early childhood caries (7.20 vs. 6.07) (30), other studies recorded a wider range than ours in salivary pH values in children with active caries (6.20-7.90) and found no correlation between pH and caries activity (30) , (31). Therefore, the static measurement of salivary pH is of little use in assessing the risk of caries (32) because, within 20-40 minutes, saliva neutralizes pH variations caused by sugary foods and the activity of microorganisms (2) , (33), while it is the acidi cation of the pH of the bio lm in the area of the lesion that determines the generation of tooth decay and correlated with a high risk of caries (2) , (16) , (33) , (34).
There are few studies of lactic acid concentrations in children's total saliva. Fidalgo et al. 2013 (35) and Pereira et al. 2019 (27) detected a higher concentration of lactate and other organic acids (acetate and n-butyrate) in the saliva of children with caries. We found higher baseline lactic acid values than did other studies (27), probably because our patients were at are high and extreme risk of caries, lactic acid concentrations were not signi cantly reduced in any group throughout the study.
The role of trace elements present in saliva on caries remains unclear (36). Of the trace elements measured in our study, Ca, P, F, Mg, Zn and Cu are, in some way, related to tooth mineralization. However, their saliva concentrations do not always re ect the degree of tooth demineralization/remineralization, caries or its risk (29) , (37). It might be thought that concentrations of Ca and P, the main components of tooth hydroxyapatite, would be increased in saliva during tooth demineralization of the tooth, although some reports have found an inverse relationship between caries and salivary calcium (31) , (38) , (39) and phosphate (38) levels, while other studies found no relationship (39).
The application of calcium phosphate varnishes should re ect an increase in the concentrations of these two ions in children's saliva, as has been observed in vitro. Cochrane et al. 2014 (40) and Shen P et al. 2011 (19), determined in vitro release of calcium, phosphate and uoride ions in ve varnishes (MI Varnish Clinpro White Varnish, Emanel Pro, Bi uorid 5, and Duraphat) and found a greater cumulative release of the ions by MI Varnish. The calcium concentrations observed in our study did not increase signi cantly after varnish application because saliva was collected 3 months after each application and there is no cumulative in vivo effect of in vitro studies, since the varnish remains in situ only for up to 24 hours as it is eliminated by chewing, salivary ow, rubbing by the cheeks and tongue and oral hygiene (41). Phosphorus was signi cantly reduced at 12 months in the control group but not in the varnish groups.
Baseline uoride levels in all groups were around 0.05 ppm and there was a non-signi cant increase, which was higher in the MI group than in the other groups, reaching 0.0920 ± 0.0402 ppm. Initial concentrations were similar to those described by Sekhri et al. 2018 (36) in cariesfree groups and higher than those described by Dehailan et al. 2017 (42) and Rechman et al. 2018 (43). As one of the exclusion criteria was consumption of uoridated running water, the higher levels of baseline salivary uoride of study children might be due to consumption of external sources of uoride, such as bottled waters. In fact, a study that analyzed the uoride content of 20 bottled water brands marketed in the area of origin of the study children found uoride concentrations between 0.05 ppm and 0.95 ppm (44). Increases in salivary uoride seem to predict increases in uoride content of dental plaque uid, implying that both would be good indicators of intraoral levels of uoride (45).
There is no consensus on the signi cance of Cu, Zn and Mg levels in relation to caries. Brookes et al. 2003 (46)  Saliva is the main remineralizing agent and generally protects the teeth. Knowledge of its composition may help detect de ciencies in patients at high risk of caries and thus provide individualized treatments that reverse the risk (47). The parameters studied in our work did not serve to de ne a risk situation or monitor the treatment with uoride varnishes and calcium phosphate. Continuous salivary ow and the in uence of the diet, hormonal status, hydration status and anxiety levels have on salivary composition, and the short half-life of varnishes in the oral environment may have been in uencing factors. It is estimated that the half-life of CPP in plaque is 124.8 minutes and that casein is hydrolyzed by salivary bacteria in a similar time (11). Therefore, varnishes must release their ions in a relatively short time and the initial high ion load is diluted over time.
Although there is no difference in the composition of some elements between plaque and saliva (sodium, ammonium, potassium, magnesium and chlorine), the metabolic activity of bacterial plaque can vary the inorganic composition between it and saliva (50). It may be useful to measure the composition of bacterial plaque and not the composition of saliva to determine the risk of caries and their monitoring.
Controlled, randomized clinical trials often have a problem due to patient losses throughout the study. Likewise, children at high or extreme risk of caries often miss appointments due to socioeconomic or other factors.
In conclusion, the quarterly application of two calcium phosphate varnishes, MI Varnish and Clinpro White Varnish, for 12 months, did not change pH, lactic acid concentrations or most trace metals in the saliva of children at high and extreme risk of caries, and therefore were not useful in in detecting a high risk of cavities or monitoring their follow-up.

Declarations
Ethics approval and consent to participate Ethics approval The statement detailing Ethics approval is exposed in the manuscript´s 115-116 line numbers.

Consent to participate
Written informed consent to participate in the study was obtained from participant´s parents or legal guardians and the statement to this effect appears in the line numbers 128-129 of the manuscript.

Availability of data and materials
All data generated or analyzed during this study are available within the article and its supplementary materials or from the corresponding author, APP, upon reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was nanced solely by the University of Murcia.