In vitro study on the preventive effect of children’s toothpastes on erosive tooth wear of primary bovine enamel and dentin

This in vitro study aimed to analyse the effect of various children’s toothpastes on erosive tooth wear of primary bovine enamel and dentin. Enamel and dentin specimens (n = 12) were cyclically eroded (6 × 60 s/d, citric acid, pH 2.4) and brushed (2 × 15 s/d, 2 N) over 5 days. Nine fluoride toothpastes (500 to 1450 ppm) and eight toothpastes containing no fluoride or other active ingredients (hydroxyapatite and/or xylitol) were tested. Unbrushed specimens served as control. Tissue loss was quantified using widefield confocal microscopy and statistically analysed using two-way and one-way ANOVAs followed by Scheffe’s (enamel) or Tamhane’s (dentin) post-hoc tests (p < 0.05). Only one fluoride toothpaste (1400 ppm) was able to reduce erosive wear of enamel significantly by 15% compared to the control (padj. = 0.002). All fluoridated toothpastes reduced dentin surface loss significantly by 32 to 69% compared to the control (padj. ≤ 0.001), while fluoride-free toothpastes were unable to reduce dentin loss significantly (padj. ≥ 0.971). Most fluoridated toothpastes caused less erosive wear compared to fluoride-free toothpastes. Children toothpastes containing fluoride were more effective in reducing erosive wear compared to toothpastes containing no fluoride or other active ingredients.

higher fluoride concentration (1100 ppm, 5000 ppm) were effective in reducing dentin loss 16 . Due to the different composition of primary compared to permanent dentin, it is also conceivable that toothpastes with lower fluoride concentration can reduce erosive wear in primary dentin. Finally, toothpastes with active ingredients other than fluorides, e.g. hydroxyapatite, were hardly analysed so far, although they are increasingly available on the market 17,18 .
Therefore, the aim of the present study was to analyse the effect of children's toothpastes with different active ingredients on erosive tooth wear of bovine primary enamel and dentin. The null hypotheses were (1) that erosive tooth wear in primary bovine enamel does not differ between the groups (brushing with the different toothpastes, erosion only) and (2) that erosive tooth wear in primary bovine dentin does not differ between the groups (brushing with the different toothpastes, erosion only).
For repositioning purposes during surface analyses (widefield confocal microscopy), identification marks (holes) were applied on the acrylic resin surfaces. Baseline measurements were recorded. Then, the outer thirds of the specimens, including the identification marks, were covered with tape (Leukoflex Pflaster; BSN medical, Hamburg, Germany) to protect the reference areas throughout the experiment. The tape was removed for the final analysis, to perform the superimposition of images captured by widefield confocal microscopy.
Prior to the experiment, the specimens were stored in distilled water. Bovine enamel and dentin specimens were randomly assigned to n = 12 specimens per group. Study design. Bovine specimens were subjected to a 5-day erosive cycling experiment consisting of six erosions and two brushing treatments per day. Erosion was performed with citric acid (Roth, Karlsruhe, Germany, pH 2.4, 1 ml) for 60 s followed by storage in artificial saliva 19,20 for 60 min. Artificial saliva was prepared based on previous studies 19,20 and contained 0.33 g KH 2 PO 4 , 0.34 g Na 2 HPO 4 , 1.27 g KCl, 0.16 g NaSCN, 0.58 g NaCl, 0.225 g CaCl 2 · 2 H 2 O, 0.16 g NH 4 Cl, 0.2 g urea, 0.03 g glucose, 0.002 g ascorbic acid mixed in 1000 mL distilled water.
One hour before the first and one hour after the last erosive treatment, the specimens were brushed in an automatic brushing machine (Willytec, Gräfelfing, Germany) with children's toothbrushes (Dr. Best children toothbrush up to 7 years; GlaxoSmithKline Consumer Healthcare, Munich, Germany). The toothbrushes were applied with linear reciprocating movements (80/min) at 2 N brushing force. The toothpaste slurry was prepared by mixing the respective toothpaste (Table 1) with water (1:2) 21 . Specimens were immersed in 3 ml of the toothpaste slurry for a total of 120 s, and were brushed for 15 s (that means 20 strokes) during this time period. Overnight the specimens were stored in artificial saliva.
Analysis of enamel and dentin wear. The surface profile of each specimen was measured before and after the erosion-abrasion cycle with widefield confocal microscopy (SmartProof 5 and ZEN Smartproof HF2 1.0; Zeiss, Oberkochen, Germany). Oriented on the reference markers an 1125 × 4500 µm 2 image was recorded with a 10× magnification (C Epiplan-Apochromat 10×/0.4 objective lens, Zeiss, Oberkochen, Germany). The used settings were a fast scan with 150-200 z-levels in mode HDR and low resolution.
Image processing and evaluation was performed with a surface-metrology software (ConfoMap 7.4.8076; Zeiss, Oberkochen, Germany). Before the evaluation the underground was leveled over rotation using a leastsquare method, furthermore background noise was reduced by removing outliers. Oriented on the reference markers, the surface profile was determined before and after the erosion-abrasion cycle on a specific position of the specimens. Over a superimposition of the both surface profiles, the surface loss was determined from the average height over 0.5 mm in the middle of the specimen. For each sample, a three-fold determination was carried out. Statistical analysis. Specimen size calculation was based on a preliminary study which used a similar design (bovine enamel erosion with hydrochloric acid (pH 2.3, 0.005 M, 60 s, 6 times daily, 5 days)). Considering that enamel loss amounted to 9.2 ± 1.5 μm and a reduction of 20% is defined as clinically relevant, a group size of n = 11 (α = 0.05, 1 − β = 0.80) was calculated (http:// clinc alc. com/ stats/ sampl esize. aspx). To consider a potential drop out of one specimen per group, specimen size per group was set at n = 12.
Statistical analysis was performed using the software SPSS Statistics for Mac 28.0.0.0 (IBM; Armonk, NY, USA). Enamel and dentin loss (mean ± standard deviation) was calculated for all groups. Normal distribution of the data was tested using Shapiro-Wilk test. As most groups presented a normal distribution, results were analysed by two-way ANOVA using the substrate (enamel or dentin) and the children's toothpastes as factors. Subsequently, one-way ANOVAs were separately performed for both substrates and followed by Scheffe's (enamel, www.nature.com/scientificreports/ homogeneous variances) or Tamhane's (dentin, nonhomogeneous variances) post-hoc tests (p < 0.05) were performed to assess potential differences in surface loss between children's toothpastes.

Discussion
In primary bovine enamel, we found significant differences between the groups as toothbrushing with ELMEX Junior reduced and toothbrushing with Weleda Kinder-Zahngel, Splat Baby, Splat Kids and Splat Junior increased tooth wear compared to control. Moreover, six out of nine fluoride toothpastes caused less wear than fluoridefree toothpastes. In primary bovine dentin, all fluoridated toothpastes, but none of the fluoride-free toothpastes reduced surface loss significantly compared to the control. All fluoride toothpastes caused less wear compared to the fluoride-free toothpastes (except for KINDER KAREX and Splat Junior). Therefore, both null hypotheses were rejected.
Bovine calves' teeth rather than human primary teeth were used as a large amount of sound primary enamel and dentin specimens was necessary for the experiment and human primary teeth are usually extracted due to extensive carious lesions. Moreover, the structural variability of human teeth (age, fluoridation, diet) is probably larger compared to bovine teeth. Erosion-abrasion experiments have shown that enamel loss is higher in specimens from calves' teeth compared to human primary teeth 22 , while the opposite was observed for dentin specimens 23 . Nevertheless, it seems appropriate to use bovine rather than human dental hard tissue when relative differences compared to the respective control are of interest 24 .
The erosion and abrasion parameters applied in this study followed guidelines for erosion-abrasion experiments aiming to simulate the everyday situation in children suffering from erosion as closely as possible 24 : Citric acid at a pH 2.4 was applied for 60 s per cycle, thus the duration time did not exceed 2 min/cycle, which is recommended as limit for the simulation of extraoral erosion. To simulate clinical conditions, a time delay between erosion and toothbrushing abrasion and vice versa of 1 h was considered by storing the specimens in artificial saliva. The rehardening potential of the artificial saliva used in this study is comparable to human saliva 20 . Toothbrushing was performed twice daily for 15 s (20 strokes) to reflect daily toothbrushing frequency and duration of most children [25][26][27][28] , and the contact time between the toothbrush and a single tooth during toothbrushing 24 . Specimens were immersed in the respective toothpaste slurry for additional 105 s (total including brushing: 120 s), to reflect the contact time of the toothpaste slurry with all teeth during toothbrushing. Toothpaste slurry was prepared at a ratio of 1:2 with water as diluent. The proportion of 1:2 to 1:4 is most commonly used for preparing toothpaste slurries in vitro, with water or artificial saliva as diluent 21,29 . However, at a ratio of 1:2, brushing treatment of eroded enamel was not different between fluoridated toothpastes diluted with water or artificial saliva 21 . Toothbrushing force was adjusted to 2 N, as both children and caregivers brushing the teeth of children apply about 2 N force 30,31 . Erosive tooth wear was analysed by widefield confocal microscopy. Confocal scanning microscopy is increasingly used in the last decade to analyse surface texture parameters and quantify surface loss with a high sensitivity 32,33 .
The present in vitro study has several strengths, such as the broad range of children's toothpastes under analysis with different active ingredients as well as the standardized experimental conditions and quantitative surface analysis in accordance with guidelines for erosion-abrasion experiments. On the other hand, the in vitro setting comes also along with limitations: One shortcoming is that the interaction with salivary components, especially the salivary pellicle, cannot be mimicked adequately in vitro 20 . Several studies have shown that oral conditions enhance the effects of active ingredients of toothpastes and mouthrinses or vice versa that the application of certain oral care products leads to an engineering of the salivary pellicle increasing its erosion protective capacity [34][35][36] . However, as relative rather than absolute values are of interest, it seems reasonable to use an in vitro design with artificial saliva to roughly assess the erosion-protective capacity of oral care products. Moreover, as commercially available toothpastes were analysed, we could not control for specific toothpaste characteristics, such as wettability, pH, size and mount of abrasive particles 37,38 .
Interestingly, toothbrushing did not lead to a significant increase of surface loss compared to the erosion only groups (except for Splat Baby, Splat Kids and Splat Junior in enamel), although previous studies demonstrated that eroded primary and permanent enamel 22,[39][40][41] and dentin 23,42 as well as enamel presenting initial caries 43  www.nature.com/scientificreports/ and even counteracted by the effects of fluorides 37,38,[44][45][46] . In the present study, each toothbrushing was performed for 15 s (equal to 20 brushing strokes) leading to an incomplete removal of the erosively softened surface layer resulting to less than 100 nm enamel loss loss 47 . In dentin, the effectiveness of active ingredients is depending on the presence of the organic matrix 48,49 , which is hardly affected by short-term brushing of 15 s 50 . The selected toothpastes contained either no active ingredient or fluoride in concentrations between 500 and 1450 ppm, hydroxyapatite, or xylitol. In primary bovine enamel, only one fluoride toothpaste containing 1400 ppm amine fluoride was able to reduce erosive surface loss significantly by about 15%, while all other fluoride toothpastes reduced enamel loss slightly, but not significantly. This result is in line with previous studies showing that toothpastes containing monovalent fluoride compounds (sodium fluoride, amine fluoride) in concentrations of 1000 to 1450 ppm fluoride have only a minor effect on erosion of permanent enamel, when tested in an erosion-abrasion model similar to the present study 44,45 . However, in the present study most of the fluoridated toothpastes caused significantly less enamel wear compared to fluoride-free toothpastes. This is also supported by the results of Ganss et al. 44,45 demonstrating no significant erosion-protective effect of brushing with fluoride-free hydroxyapatite containing toothpastes.
Generally, fluoridated toothpastes showed a higher erosion-protective effect in primary bovine dentin compared to enamel. The protective effect is in the range of previous studies analysing the effect of brushing with sodium fluoride or amine fluoride containing toothpastes on erosive tooth wear in permanent dentin 46,51 .
Interestingly, no clear dose response effect was observed for the fluoride containing toothpastes, probably as the overall fluoride concentration was too low to exhibit differences in the erosion-protective potential. Toothpastes of the same brand (same composition except for fluoride concentration, same abrasivity) showed no or only slight differences between different fluoride concentrations. Nevertheless, it has to be emphasized that even toothpastes with low fluoride concentration were effective in reducing erosive dentin loss, which might be of particular relevance for children suffering from severe erosive tooth wear already affecting dentin. Differences between toothpastes with the same fluoride concentration, but of different brands might be related to the abrasivity (RDA-value), particle type and further physical and chemical factors of the toothpaste, which can-at least partly-affect the protective properties of active ingredients 37,38,52 .
Five of the fluoride-free toothpastes analysed in the present study contained hydroxyapatite, partly in combination with xylitol. Previous studies on a potential erosion-protective effect of fluoride-free hydroxyapatite toothpastes showed that these toothpastes were unable to prevent erosive loss of permanent enamel and dentin [44][45][46] . Recently, it has been demonstrated that brushing (3 × 3 min/day, 15 days) with a zinc-carbonate nano-hydroxyapatite toothpaste was unable to form a newly-mineralized surface layer on enamel and dentin 53 , which then might protect the underlying surface against erosion.
Very limited information on the erosion-protective effect of xylitol-containing toothpastes is available so far 54 . Brushing of eroded enamel with a 10% xylitol toothpaste resulted in less wear compared to a placebo toothpaste and was not significantly different from a fluoridated toothpaste (1030 ppm, NaF). The authors speculated that xylitol acts as lubricant, reducing the physical impact of brushing on the eroded surface 54 . However, in the present study the toothpastes that contained only xylitol and no further active ingredients (nenedent Kinderzahncreme ohne Fluorid, Putzi Kinderzahngel) were unable to reduce erosive enamel and dentin wear compared to control and were significantly less effective compared to the fluoridated toothpastes from the same brand.

Conclusion
Within the limitations of an in vitro study, the present results indicate that fluoride-containing children's toothpastes rather than toothpastes without fluoride or with other active ingredients should be recommended for the prevention of erosive tooth wear in the primary dentition, especially when primary dentin is affected.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.