It was an experimental study conducted at the Nucleus of Identification and Analytical Quantification (NIQUA) of the Pharmacy School (Federal University of Juiz de Fora), with a duration of 10 days (from 8^th May 2017 to 18^th May 2017).

This study was developed in accordance with the Ethics Committee of the Federal University of Juiz de Fora (n° 1.345.640). Thirty intact human third molars extracted for clinical reasons and obtained from Human Teeth Bank of the Dental School (Federal University of Juiz de Fora) were selected for this study. The sample size was based on the methodology established by Lussi A et al., and Sener Y et al., [3,9]. Teeth were disinfected at room temperature with sodium hypochlorite for 24 hours [9].

In sequence, all teeth were submitted to horizontal sections at the cemento enamel junction, separating coronary and root pieces, with a metallographic cutter (Isomet^® 1000 Precision Saw, Buehler, United States) and a diamond disc (15 LC, Diamond Wafering Blade, Buehler, USA). Then, dental coronal sections were made, at the middle third, so that only one dental block for each tooth was obtained and the cuts were made in the flat area under cooling with distilled water to avoid cracks in the enamel, obtaining enamel blocks with dimensions of 4 mm×4 mm×2 mm [10,11]. The fragments were embedded in acrylic resin block with the buccal surface projected to the external environment, ground flat and polished [10,11]. The buccal surfaces were abraded using a belt sander (Politriz PL02, Teclago, Brazil) with grit sandpaper 600, 1000, 2000 and 4000, under finger pressure. Felt discs and diamond pastes (Diamond AC I and II) were used to promote the final polishing [10]. At the end, the enamel blocks were washed with distilled water (Direct-Q^®, Millipore, France) in an ultrasonic cleaning bath (USC 1400, Unique, Brazil) for five minutes and stored in plastic containers with absorbent paper, moistened with distilled water and stored in a refrigerator at 4°C.

The experimental material (stem bark) was herbal tea bags of Caesalpinia ferrea (Martius), Schinus aroeira (Vell), Uncaria tomentosa (Willd) DC., Morinda citrifolia L., and Schizolobium amazonicum F Smith available in the retail trade (Moinho Central Indústria Comércio, Belém, Pará, Brazil). Teas were prepared with water, according to the manufacturer’s instructions, kept at room temperature and used on the same day. Samples of ready to drink tea were collected directly from the product package immediately after its opening.

The measurements of the beverages pH values were made at 25°C using an electronic pH meter (B474, Micronal, São Paulo, Brazil) connected to a glass electrode. Prior to the measurement, the equipment was calibrated using standard solutions with known pH 7.0 and 4.0 (Indústria Comércio eletro-eletrônica Gehaka Ltda, São Paulo, Brazil). Afterwards, three readings of the freshly prepared drinks were obtained in order to give a mean pH measurement for each sample.

To determine titratable acidity (buffering capacity), 50 mL of each beverage was combined with 0.1 mL increments of 0.1 mol/L sodium hydroxide (NaOH). The amount of NaOH required to reach pH level of 7.0 was recorded. The titrations for each beverage were also repeated two times to obtain a mean value.

The study was conducted in an erosion remineralisation cycling model, in order to simulate the conditions in the human oral cavity. The specimens were immersed in artificial saliva {(0.12% KCl, 0.089% NaCl, 0.005% MgCl₂, 0.146% Ca₃(C₆H₅O₇)₂, preservatives (0.002% nipagin and 0.013% nipasol), 1.0% Carboxymethyl Cellulose (CMC), 3.0% sorbitol, distilled water (quantity sufficient for 100 mL)}, according Silva AF et al., at a volume of 2.5 mL/mm² of the enamel surface and were maintained under soft agitation for 30 minutes before the first erosive challenge and then washed in running deionised water [12].

The enamel blocks (n=5 specimens) were then randomly assigned into six groups and immersed in the respective treatment groups (teas): A) M. citrifolia; B) U. tomentosa; C) C. ferrea; D) S. amazonicum; E) S. aroeira; F) positive control (1% citric acid) and remained still, without stirring for two minutes at room temperature, five times a day with intervals of two hours between the five daily erosive challenges, lasted for five days. During the interval between erosive challenges, the samples remained in artificial saliva and then were washed with deionised water and lightly dried with absorbent paper. The assay was performed with five specimens and in triplicate.

Before the beginning of the erosion cycle and the sequence cycles, samples were analysed for roughness with a portable surface roughness tester (SURFTEST SJ-301, Mitutoyo, Japan). Each group was subjected to surface microhardness measurement to obtain a baseline value. The hardness value (0.05 kg/mm²) of each specimen was determined using a microhardness test (HMV-2, Shimadzu, Kyoto, Japan). The average of these values (n = 15) was used to calculate the percentage of surface hardness loss, using the following formula:

In relation to the roughness of the surface of the samples, the difference between the final roughness and the initial roughness was analysed. Both assays were performed with five specimens and in triplicate.

Statistical analyses were performed using the IBM SPSS statistics 22.0 software. The results were expressed as mean±standard error of the mean and were analysed using analysis of variance (ANOVA) followed by Tukey’s post-hoc test. The accepted significance was set at p-value <0.05.

The 1% citric acid had a pH of 2.14±0.01 and the pH of the teas ranged from 3.56 to 5.56 [Table/Fig-1]. Although, the teas were not statistically similar to the positive control group (1% citric acid), they showed values close to the critical pH for the demineralisation of dental enamel.

In reference of titratable acidity, the amount of NaOH needed to reach neutrality (pH 7.0) decreased with the following order: M. citrifolia, S. aroeira, C. ferrea, S. amazonicum, and U. tomentosa [Table/Fig-2].

All of the tested teas promoted significant loss of superficial enamel hardness (34.12 to 46.98) and there was no statistical significant difference between the results and the positive control group (1% citric acid) [Table/Fig-3]. In relation to the surface roughness of the samples, there was a statistically significant difference between 1% citric acid and S. amazonicum, S. aroeira, and C. ferrea (p-value <0.01) [Table/Fig-4].

According to Jaâfoura S et al., there are few studies that explore the relationship between tea consumption and dental erosion [13]. Consumption of teas and damage to teeth are often associated with discolouration or with the remineralisation process, but little is known about their involvement in the pathological process of dental erosion [7,13].

Several characteristics and properties of saliva play an important role in dental erosion. Erosion makes this modulation through its buffering, dilution of acids, demineralisation and training acquired film [14]. Artificial saliva was used in this study with the intention to simulate the oral environment, offsetting the limitations of ex vivo study. After soaking in teas, there was no evidence of the remineralisation of tooth enamel, which highlights the importance of knowledge of the erosive potential of consumed teas.

The enamel begins to demineralise naturally below pH 5.5, so it is necessary to avoid reaching that lower pH or, alternatively, to make hydroxyappatite more resistant to acid dissolution [13,15,16]. Once erosion has begun, there is no method to enable the regeneration of tooth structure; among other reasons, this is because the ameloblasts secreting tooth enamel disappear before teeth sprout [15]. All evaluated teas presented a pH≤5.5, which is considered critical for tooth enamel demineralisation and demonstrates erosive potential of the teas.

If the softened layer of enamel is not rehardened and the acid impact continues, progressive softening and dissolution of the consecutive layers of enamel lead to permanent loss of volume, with a softened layer on top of the remaining tissue [17]. This softened layer on the enamel surface is highly susceptible to wear from, for example, an abrasive toothbrush insult [18]. In severe long-term erosive wear, when there is frequent substantial erosive attack on the enamel surface, the volume of hard tissue permanently lost is clearly visible clinically as a defect in the tooth structure [19].

Assessing the erosive potential of beverages, Lunkes LB and Hashizume LN elucidated the variables that may be considered influential in this process, such as calcium, fluoride, and phosphorus concentrations and the type and concentration of acids present; however, the main factor in the erosion analysis is pH [7]. Measurement of the TA has also been used to evaluate the erosive potential of beverages. This method considers the type and concentration of the acid present. According to these authors, the majority of erosive beverages have weak acids and the concentration of these acids determines not only the pH value but also the buffering properties. Their analyses used corroborate with the present study, which also opted to evaluate pH and TA, but the concentrations of these ions were not measured.

Soares AK et al., emphasises that the TA of a beverage influences salivary pH more than the pH of the beverage [20]. As a consequence of the high TA, there is an increase in the time for saliva to neutralise the acid, causing differences in the erosive potential of a beverage even within the same pH range. The TA reflects the amount of a base that must be added to a beverage to raise the pH to 7.0 (neutral), and represents the erosive potential of the beverage [21]. Thus, a beverage with low TA is readily neutralised by salivary buffers, preventing prolonged dropping of buccal pH and, consequently, causing less mineral loss in dental structure [21].

According to Amoras DR et al., chemical composition of an acidic beverage clearly has an important effect on the hardness properties of enamel [22]. Jameel RA et al., reported that dental erosion induces demineralisation, thus affecting the hardness of dental tissue, which may be related to the degree of remineralisation [16]. This explains the loss of hardness due to erosion of dental enamel in the present study, in which all the analysed teas were statistically similar to 1% citric acid (45.09%, p-value >0.05). According to De Moraes MD et al., they agree that hardness analysis is a sensitive method for detecting changes in mineral density of eroded surfaces/abrasive lesions on a substrate enamel, but not on dentin [23].

According to Pereira CA et al., most beverages available on the market have high acidity and are able to demineralise enamel and dentin [24]. These demineralised dental structures are more susceptible to wear and abrasion by brushing, thereby accelerating the loss of dental surface structure, which can be assessed indirectly through surface roughness. The surface is considered rough if it is characterised by the presence of peaks and valleys of high amplitude and reduced ripple. The surface roughness value considered critical for the retention and adhesion of microorganisms is 0.2 μm. This finding corroborated that of the present study, in which all the teas were statistically similar to citric acid (increase in surface roughness after erosive cycles), except S. amazonicum, S. aroeira, and C. ferrea teas. Further, roughness directly interferes with the aesthetic properties of the tooth and restorative materials, such as gloss and surface smoothness, colour change and staining, dental biofilm accumulation and ultimately secondary caries lesion formation. Therefore, researches that measure the effect of acidic beverages on the structure of dental enamel are of great value. According to Barac R et al., apart from the chemical aspects of erosion, physical characteristics of the enamel surface were evaluated to indicate the loss of tissue due to exposure to soft drinks with low pH values [5]. These characteristics can be evaluated by measurement of roughness parameters along a line by contact stylus surface profilometry.

The technical difficulty of standardising the sample cuts and the absence of other factors that could influence dental erosion. In spite of this, it is suggested that it provides valuable insight into the erosive effects of tested teas, which are widely consumed in Brazil.

The combined analysis of the results allows us to conclude that all evaluated teas presented a pH ≤5.5, considered critical for demineralisation of tooth enamel and the presence of ions capable of interacting with the dental surface and causing loss of dental tissue. The mean roughness values of M. citrifolia and U. tomentosa teas were similar to those of citric acid. As many of the teas tested were at least as erosive as 1% citric acid the potential for harm is obvious. Thus, there is a need to be careful regarding consuming such beverages.