2.1

Saliva collection

Using the spitting method, un-stimulated saliva was collected by asking subjects to spit into a pre- weighed container over a period of 5 min. They were then asked to chew on a piece of paraffin gum and at the same time spit into a separate container over a period of 5 min. Flow rates were calculated and recorded in ml/min.

2.2

Sample preparation

The crowns of twelve human upper central incisors were gently abraded and pumiced using 1200 – grit sandpaper (Water Sandpaper, 151 Products Limited, Manchester, UK) to remove extrinsic stains and other residues and to obtain a homogenous surface. Each crown was sectioned in a labio- palatal direction (n = 12 labial, 12 palatal specimens) using water – cooled diamond wire (Well type 2400, Walter EBNER, Le Locle, Switzerland). The labial and palatal surfaces were further sectioned mesio- distally to produce 2 specimens from each surface (n = 24 labial, 24 palatal specimens).

The incisal and cervical thirds of each specimen were discarded and the remaining surface was reshaped. Therefore, each incisor was sectioned both labio- palatally and mesio-distally to give four 4 × 4 × 1.5 mm specimens per tooth ( Fig. 1 ). The dimensions of all specimens were standardized using Vernier calliper then they were all checked under a USB microscopic light (Veho™, Discovery, UK) to ensure dentine removal.

Schematic representation illustrating crown sectioning method.

2.3

Specimen allocation and insertion

A total of 24 labial and 24 palatal enamel specimens were randomly allocated to a 12 upper removable appliances. This was performed in such a way that 2 labial and 2 palatal enamel specimens were randomly allocated to their corresponding sites on each appliance (n = 4) so that one labial and one palatal enamel specimen were positioned in the anterior and another 2 in the posterior region ( Fig. 2 ).

Upper removable appliance with all surfaces of the specimens exposed to abrasion from the tongue. L = labial cut, P = palatal cut.

Fig. 2 shows the surface and position of each enamel specimen on the appliance. Each appliance had 4 specimens but each specimen was subjected to either an erosion only regimen (palate- facing side) or an erosion and abrasion regimen (tongue – facing side). Therefore, a total of 96 surfaces were assessed.

2.4

Design of the removable appliance

Twelve upper removable acrylic base plates, each had 2 retention clasps on both maxillary first and second premolars were constructed ( Fig. 1 ). Four 1.0 × 0.7 cm slots; 2 opposite the central incisors and 2 next to the first molar teeth were cut through each base plate to accommodate 4 jigs of the same size. On one side of each slot, a hinge joint made of [- shaped 0.5 mm orthodontic wire was embedded in the acrylic resin into which an acrylic jig holding 1 enamel specimen was attached. In order to easily identify and quickly assign each specimen to its corresponding slot on the appliance, each acrylic jig had a groove on the side representing the number of each specimen e.g. specimen 1 had one groove; specimen 2 had two grooves, etc. The hinge type joint facilitated the removal and reinsertion of each specimen in its corresponding slot on each appliance.

During manufacture of the acrylic base plates a single layer of wax was placed in the slots against the palate to standardise the gap and in effect countersink the palate facing surfaces of each enamel specimen. Specimen surfaces facing the palatal mucosa were thus slightly recessed, by approx. 1 mm inside the appliance, to avoid any possible friction and thus abrasion by the opposing palatal mucosa. The surfaces were, however, still accessible to the erosive challenge allowing each specimen to act as its own control ( Fig. 3 ). Each appliance then had 4 enamel specimens; 2 from labial and 2 from palatal enamel.

A palatal view of a removable appliance showing surfaces facing the palatal mucosa and not subjected to abrasive tongue forces but only to the in vitro acid erosion from Orange Juice (red arrow). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Differences of the mean fluorescence loss for the 48 specimens/ 96 surfaces of specimens during the 4 week study period. Note: differences were calculated by subtracting %∆F of the tongue- facing surface from that of the palate- facing surface. Specimen position in the appliance is shown whilst its section from the tooth is denoted as either L for labial or P for palatal.

%ΔF for the labial and palatal sections during the 4-week study period. No significant statistical difference between labial and palatal surfaces was found at the end of each week of the study (p > 0.05).

2.5

Sterilisation

The 12 upper removable appliances with the jigs inserted were sterilised by exposing them to Gamma irradiation using a cobalt – 60 source with particle energy of 0.315 MeV and irradiation cycle of 3.4 Gy / min at a source – to – specimen distance of 100 cm and field size of 15 × 15 cm for 2 days making a total overall dose of approximately 25 kGy .

2.6

In situ study

On each day of the study, volunteers were asked to wear the intra- oral appliance and its retained enamel specimens between 09.30 and 1700 h except for 1 h over lunch- time and two 30 min coffee breaks. When not in place inside the mouth, appliances were taken from the volunteers, jigs containing the specimens detached and immersed, for 5 min, in 50 ml of orange juice (Tesco value from concentrate, Tesco, UK) before each re – insertion. The mean pH of the orange juice was 3.85. The Adams cribs and collets around the teeth ensured seating in the same position after each re-insertion.

The first 90 min intraoral insertion of the appliances, however, was performed without exposure to the erosive challenge to allow the formation of a salivary pellicle layer, mimicking the natural situation found inside the mouth.

Volunteers were not allowed to consume food products nor beverages or practice any oral hygiene measure while wearing the appliance. Overnight and at weekends, the appliances and their specimens were stored in humid containers at room temperature until next use.

2.7

Erosion/erosion and abrasion lesion measurement

Quantitative light – induced fluorescence images (QLF, version 2.00c, Inspektor Research Systems BV, Amsterdam, Netherlands) were taken and analysed using an external reference composite jig to which the fluorescing intensity of tooth surfaces could be referenced ( Fig. 6 ). QLF measurements at 5% threshold were used. Non-contact light profilometry scans (NCLP, Proscan 2000, Scantron, Taunton, UK) were performed by marking three fixed points on each acrylic jig and using each as a reference ( Fig. 9 ).

An example of a tongue- facing surface QLF final image analysis, for one of the anterior specimens, at the end of each week of the study period. QLF demonstrates severity of enamel surface loss in false colour. Red fluorescence in the middle of (a) and (b) images denotes mean fluorescence loss at 25% threshold. As tongue abrasion continued, %∆F increases reaching 45% threshold level (yellow) by the end of week 4. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Comparison of the differences in the MSH (MSHD(t-p)) for the 48 specimens/96 surfaces during the 4 week study period. Note: positive values indicate increased DMSH on the tongue- facing surfaces; negative values indicate increased DMSH on the palate- facing surfaces. Specimen position in the appliance is shown whilst its section from the tooth is denoted as either L for labial or P for palatal.

Mean step height for the surfaces from labial and palatal sections irrespective of which tissue they faced (n = 96 surfaces) over the 4 weeks period. There was no significant statistical difference between labial and palatal specimens at the end of each week of the study period (p > 0.05).

Examples of profilometry scans of the tongue- facing surface ( top ) and palate- facing surface ( bottom ) of a specimen showing the changes in the step height from baseline ( left ) to week 4 ( right ). Note: the three markings on each image were carved on each acrylic jig to standardize specimen orientation after each cycle.

QLF images and NCLP scans were taken for the entire tongue- facing and palatal- facing surfaces of each specimen, at baseline and at the end of each week throughout the 4-week study period. Measurements were made, however, on the centre of each specimen. Data reported were ΔF (% fluorescence loss of enamel lesion created compared to sound enamel). NCLP mean step height was systematically measured across four mid – points on the sides of each specimen to give four measures of step height (loss of depth). A mean value for the step height for each specimen was then calculated.

2.8

Statistical analysis

Means were calculated for saliva pH and flow rate. The data obtained were statistically analysed using SPSS V20 (SPSS Inc., Chicago, USA). Changes within each specimen were recorded from both labial (erosion and abrasion) and palatal (erosion only) surfaces at baseline and at the end of each week. As each specimen acted as its own control, the analysis of QLF and NCLP data was performed on the mean differences between the two different lesions rather than on absolute values. Those differences, for all measurements, were calculated by subtracting the mean fluorescence loss and mean step height values on the tongue – facing surfaces from those on the palate – facing surfaces. A one – way ANOVA and paired t -test were used for comparisons between mean differences throughout the study period. A p- value < 0.05 was considered statistically significant.