Head shape and abnormal appearance of tympanic membranes.

We examined 187 males to evaluate the hypothesis that short-headed persons more often have otitis media than long-headed persons. The subjects were seen as part of routine health screening before attending recreational summer camp. The photographic appearance of each tympanic membrane was judged on two separate occasions by an otolaryngologist who categorized them as "normal", "abnormal", or "cannot determine". The repeatability of the categorizations was fair: kappa = 0.44. Only the tympanic membranes categorized identically twice were used for data analysis. The left-right symmetry of the tympanic membrane categorizations was fair: correlation coefficient phi = 0.42. Subjects categorized as having bilaterally normal tympanic membranes were considered to have not had previous otitis media (N = 95). In contrast, subjects categorized as having at least one abnormal tympanic membrane were considered to have had previous otitis media (N = 13). Head length and width, measured by calipers, and the cephalic index (width divided by length) x 100 were determined for each person. On the average, the cephalic index of the normal subjects was lower than that for the subjects with otitis media (t = 3.06, p less than 0.005). These data support the suggestions of Pautow (1925) and Worley et al. (1987): dolichocephalic persons have otitis media less often than do brachycephalic persons. Though this association is not useful in clinical care of the individual patient, it may be considered a weak external manifestation of the different cranial base and eustachian tube anatomy found in persons with otitis media.