RESUMO
The analysis of microwear patterns, including scratch types and widths, has enabled reconstruction of the dietary habits and lifestyles of prehistoric and modern humans. The aim of this in vitro study was to determine whether an assessment of microwear features of experimental scratches placed on enamel, perpendicularly to the direction of grinding, could predict the grinding direction. Experimental scratches were placed using a scalpel blade on standardised wear facets that had been prepared by wearing opposing enamel surfaces in an electromechanical tooth wear machine. These control 'baseline' facets (with unworn experimental scratches) were subjected to 50 wear cycles, so that differential microwear could be observed on the leading and trailing edges of the 'final' facets. In Group 1 (n=28), the 'footprint' microwear patterns corresponding to the known grinding direction of specimens in the tooth wear machine were identified. Then, they were used to predict the direction of tooth grinding blindly in the same sample after a 2-week intermission period. To avoid overfitting the predictive model, its sensitivity was also cross-validated in a new sample (Group 2, n=14). A crescent-shaped characteristic observed in most experimental scratches matched the grinding direction on all occasions. The best predictor of the direction of grinding was a combined assessment of the leading edge microwear pattern and the crescent characteristic (82.1% in Group 1 and 92.9% in Group 2). In conclusion, a simple scratch test can determine the direction of tooth grinding with high reliability, although further improvement in sensitivity is desirable.