Material and methods

Independent-group t test analysis estimated that 32 subjects per group were required to detect a difference of 2 mm in overjet among the various groups. This was based on an alpha significance level of 0.05 with 95% power. The standard deviation for the sample size was calculated as 1.9 mm based on data from the National Health and Nutrition Examination Survey III, which included detailed data on overjet across a large population of white subjects.

Ethical approval was obtained from the University of Birmingham Research and Ethics Committee and the Royal Northern College of Music Research and Ethics Committee in the United Kingdom. All subjects were treated according to the Declaration of Helsinki (1964) and the British Psychological Society’s code of ethics and conduct (2006). Signed informed consent was obtained from each subject who participated in the study.

Various well-established and reputable professional orchestras, jazz bands, and music colleges were contacted. The subjects comprised wind instrument players of classes A and B (Strayer’s classifications), and the control group comprised string instrument and percussion players, recruited from the same organizations.

The subjects were therefore separated into 4 groups: (1) players of large-mouthpiece brass instruments, such as the tuba and trombone; (2) players of small-mouthpiece brass instruments, such as the trumpet and French horn; (3) players of single-reed instruments such as the clarinet and saxophone; and (4) players of string and percussion instruments (control group).

The subjects were selected from the following organizations: Royal Philharmonic Orchestra, City of Birmingham Symphony Orchestra, London Philharmonic Orchestra, Lucerne Symphony Orchestra, Oxford Philomusica, Royal Liverpool Philharmonic Orchestra, BBC Big Band, Welsh National Opera Orchestra, Halle, BBC Philharmonic Orchestra, Opera North, National Saxophone Choir, Clarinet and Saxophone Society, Trinity College of Music, BBC Scottish Orchestra, Royal National Scottish Orchestra, Royal Scottish Academy of Music and Drama, BBC Concert Orchestra, and Royal Welsh College of Music and Drama. Furthermore, 2 organizations, the East London Clarinet Choir and the Notebenders Jazz Club, were visited, but no musician satisfied the selection criteria.

The included subjects were white men and women who were professional wind instrument players, practiced on average at least 3 hours daily, and for at least the last 4 years, and started playing a wind instrument before they were 14 years old. The control group comprised professional string and percussion players from the same orchestras and organizations as the wind instrument players.

Wind instrument players who played more than 1 class of wind instruments either professionally or recreationally, with the exception of single-reed players who also played the flute or the piccolo, were excluded. String and percussion players who played or used to play a wind instrument recreationally were excluded.

In terms of their dental status, we excluded subjects (wind, string, and percussion players) who had previously undergone orthodontic treatment; had extractions of permanent teeth other than second and third molars; had retained deciduous teeth or supernumerary teeth, crowns on permanent teeth other than first, second, and third molars; had restorations on incisors and canines that extended over 2 surfaces; or had pathology, including periodontal disease, previous fractures of the maxilla or the mandible, and dental cysts. Subjects who admitted to a digit-sucking habit that persisted until the age of at least 10 years and subjects who smoked a pipe were also excluded.

The musicians who satisfied the criteria on the basis of written questionnaires had a brief examination of their teeth. Those who were eligible to participate on the basis of the inclusion and exclusion criteria had impressions taken at their practice venues. Study casts were made and coded to preserve anonymity and ensure that there was no observer bias during their assessment.

Recruitment of participants stopped as soon as 32 musicians were obtained for each group, as dictated by the sample size calculation. A total of 170 musicians participated in the study. They comprised 32 subjects in the large-cup brass group, 42 subjects in the small-cup brass group, 37 subjects in the single-reed group, and 59 subjects in the control group. Inclusion of more than 32 subjects in the first 3 groups increased the statistical power of the study.

The following features were assessed on the study casts: (1) overjet (mm), (2) overbite (mm), (3) British Standards Institute incisor relationship, (4) maxillary and mandibular intermolar widths (mm), (5) crowding in the maxillary and mandibular labial segments (mm), (6) Little’s irregularity index (mm), and (7) crossbites, with reference to the maxillary first permanent molars.

All measurements were made by 1 examiner (E.G.). Digital calipers were used to measure linear measurements such as the overbite, intermolar width, and crowding. The digital calipers were calibrated to ensure accuracy on every 12 successive study casts.

Incisor relationship was assessed according to the British Standards Institute incisor classification. This classifies incisor relationships as follows.

Class I, the mandibular incisor edges occlude with or lie immediately below the cingulum plateau of the maxillary central incisors.

Class II, the mandibular incisor edges lie posterior to the cingulum plateau of the maxillary incisors. There are 2 divisions. In Division 1, there is an increase in overjet and the maxillary central incisors are usually proclined. In Division 2, the maxillary central incisors are retroclined. The overjet is usually minimal but might be increased.

Class III, the mandibular incisor edges lie anterior to the cingulum plateau of the maxillary incisors. The overjet is reduced or reversed.

To define the intermolar widths, the mesiobuccal cusps of the maxillary and mandibular first molars were taken as reference points. Little’s irregularity index, which is the sum of the contact point displacements in the 6 anterior teeth, was measured as described by Little.

On each pair of casts, the presence or absence of a transverse discrepancy of tooth relationship (crossbite) was scored as follows.

• 1.

  Absence of crossbite was defined when the palatal cusps of all maxillary molars occluded to the midpoint of the central fossa of the mandibular molars.

• 2.

  Presence of lingual crossbites or lingual crossbite tendencies: a lingual crossbite tendency was defined when the palatal cusp of the maxillary first permanent molar occluded lingually to the central fossa of the mandibular first permanent molar. Lingual crossbite was defined when there was cusp-to-cusp contact or more.

• 3.

  Presence of buccal crossbites or buccal crossbite tendencies: a buccal crossbite tendency was defined when the palatal cusp of the maxillary first permanent molar occluded buccally to the central fossa of the mandibular first permanent molar. Buccal crossbite was defined when the palatal cusp of the maxillary first permanent molar occluded on the buccal cusps of the mandibular first permanent molar.

The descriptions and definitions of crossbites of American orthodontists are the opposite of the description of the British Standards Institute classification that defines a buccal crossbite as a transverse discrepancy in tooth relationship where the buccal cusps of the mandibular teeth occlude laterally to the buccal cusps of the maxillary teeth.

Before taking the measurements, the reproducibility of the measurements was confirmed with a matched pairs t test by remeasuring 20 randomly selected study casts 9 days later. This confirmed that there was no significant difference between initial and subsequent measurements.

One-way analysis of variance (ANOVA) was used to compare the means of interval variables among the groups: overjet, overbite, crowding, Little’s irregularity index, and intermolar widths. The chi-square test was used to assess whether there was a difference in categorical variables among the groups: the proportions of subjects with various incisor classifications and crossbites. A retrospective sample size calculation, with the standard deviation calculated from the data of the subjects of this study, confirmed that there was 83% to 98% power for the analysis and comparison of all the other occlusal features.