Conservative scoring and exclusion of the phenomenon of interest in linear enamel hypoplasia studies

Conservative criteria are generally employed to select only "sufficiently large" linear enamel defects to be scored as linear enamel hypopasias (LEHs), a practice that excludes enamel defects below arbitrarily selected size thresholds. A question arising is, are such practices likely to exclude small, yet authentic(i.e., physiological stress induced) LEH? The permanent maxillary and mandibular first and second incisors, canines, and first premolars on dental casts (n = 532) form a contemporary Canadian farming commune were scored for LEH, with linear defects being categorized by both breadth and width to yeild the scores A (smallest), B, C, and D (largest). Relative frequencies were calculated as the percent prevalence of each score per tooth divided by the highest percent prevalence exhibited by any tooth in the same dental arch. The intertooth patters of the relative frequencies were tested for concordance. Secondly, assuming that different researchers employ different scoring criteria, tests for concordance of mean n of LEHs/tooth were conducted using data from four different reports as a second test for intertooth concordance of various-sized defects. Substantial concordance was obseved between A, B, C, and D (Kandall's coefficient of concordance, W = .65, P <.005) and in the interstudy comparison (W = .67, P <.01). Because it is unlikely that defects of varying sizes would have similar intertooth distributions if they did not have similar etiologies, it may be inferred that all of the defects observed, even the very small defects, A, are variants of the same phenomenon, LEH. © 1995 Wiley-Liss, Inc.