Performing artists in general and instrumentalists in particular spend many years training from a young age to achieve high levels of skill and artistry, and this is likely to have an effect on their body. As elite athletes of the hands, prevention of injury and pain is crucial as its presence decreases performance quality and causes unnecessary time loss.
In recent years, our medical knowledge of hypermobility syndrome has increased, but little is known about injury prevalence in musicians with abnormally flexible joints1 and its effects.2 Although some studies have found hypermobility more frequently in performing artists than in the general population,3,4 few studies of adequate size have been done, none in flautists.5
On the other hand, studies of hand and arm symptoms among professional flautists and students have been conducted recently,6–8 and the impact of hypermobility in the hand in musicians has been a further area of recent interestfs.9 Any relation between proprioception and hypermobility in the hand has not previously been reported in musicians. An earlier study on a normal population found that those with higher scores on the Beighton Scale (≥4/9) had significantly impaired proprioception, especially at hypermobile joints10; it is still unclear if this disturbance is a cause or effect of hypermobility, but patients with hypermobility syndrome tend to present with a range of complaints due to the effects of joint instability such as muscle and joint pain.
The present study was designed to investigate the relationship between hypermobility and flute playing, to study the impact of the amount of practice and the years of training, as well as the ergonomics of the flute itself, and focused on how the presence of joint laxity might play a role in the hand position, with any change in proprioception that might subsequently occur.
SUBJECTS AND METHODS
Participants were flautists of different backgrounds: students in London music training institutions, professional flute players, and semiprofessionals/playing as a hobby, who had played for a minimum of 10 years. A recruitment advert was distributed by e-mail to all eligible students and teachers of the Royal College of Music, the Guildhall School of Music and Drama, and the Trinity Laban College of Music and Dance, as well as to other music backgrounds in order to recruit all types of flute players. A total of 27 flute players expressed interest in participating and were provided with the participant information statement, of whom 20 (4 male and 16 female subjects) subsequently completed the study.
All testing occurred in a single session with each participant, and the sessions took place both at the Trinity Laban Conservatoire of Music and the British Association of Performing Arts Medicine clinic. Before performance, all participants were familiarized with the equipment and were able to ask any question in relation to its operation. All flautists brought and played their own instruments.
All participants were asked to fill out a short questionnaire in order to gain some background information and to perform the 5 maneuvers of the 9-point Beighton Scale (Table 1) to gain some indication of general joint laxity. Participants were classified as hypermobile if they could perform at least 4 of the 9 diagnostic maneuvers of the Beighton Scale listed. The rest of participants (those who scored <4 points) were used as control subjects, age matching them as closely as possible to the hypermobile subjects.
Objective data were also recorded. The passive hyperextension range of motion of the 3 joints most specifically involved in flute playing (the interphalangeal joint [IP] joint of the thumb, the metacarpophalangeal joint [MCP] joint of the index, and the distal IP (DIP) joint of the little finger) of both hands was measured with the Leeds Hyperextensometer11,12 for the MCP joint (Fig. 1) and a 66fit–12″ goniometer for the IP joints (Fig. 2). The Leeds proprioceptometer13 was used to evaluate proprioception.
The hyperextensometer is a device that measures the passive extension at the MCP joint of the finger in response to a preset torque. To quantify finger hyperextension, the hand is placed flat on the base plate of the hyperextensometer. The index, or another finger, is strapped to the rotating arm and the pointer set at zero. By rotating the knob, a preset torque (2.6 kg/cm) is applied to the finger. A slipping clutch ensures that a greater force cannot be applied. The resulting angle of hyperextension is recorded by the excursion of the pointer along the dial.
The proprioceptometer (Figs. 3 and 4) was used to evaluate proprioception. Briefly, it is an apparatus that isolates the index finger, allowing full flexion and extension of the MCP of that digit, while splinting the distal interphalangeal and proximal IPs into extension. The remaining fingers and hand are also held underneath.
The experimental protocol consisted of passive displacement of the participant’s hidden index finger from a starting position (0 degrees) to 10 given positions in either flexion (10, 25, 30, 35, 50, and 65 degrees) or extension (−10, −30, −40, and −10 degrees) in a random order. At each of these angles, the participant was required to estimate the range through which the finger had been passively moved by moving the pointer to the position they considered the hidden finger to have reached. After each match, the equipment was zeroed to the starting position (0 degrees). The 10 angles were randomized to avoid a learning effect when testing the other hand.
The difference between the actual and estimated angles represented the matching error (in degrees), which had both magnitude and direction (positive if flexion error, negative if extension error). Both hands were assessed.
All information collected was tabulated for interpretation. For proprioception, a total single value was then calculated by adding the proprioceptive error at each of the angles of the finger at which it was measured.
All data were entered into a computer-based statistical spreadsheet for subsequent data analysis. Statistical data analysis was performed using the Statistical Package for the Social Sciences (SPSS version 17).
All data were organized, matching the data by variables: gender, hypermobility, and handedness. Normal distribution of the data was evaluated to determine whether to use parametric or nonparametric tests. The assessment of normality of data was performed with the Shapiro-Wilk test for small sample sizes (<50 samples).
For the “5th,” “thumb,” and “index” fingers, the dependent variable “gender” was normally distributed (P > 0.05), but for “hypermobility” and “handedness,” the data significantly deviated from a normal distribution (P < 0.05). Comparison of the right and left hand for each finger was therefore done using the nonparametric 2-related-samples Wilcoxon t test. Comparison between subgroups (hypermobility, gender, and handedness) was also performed using the nonparametric Kruskal-Wallis test of variance.
In the evaluation of proprioception acuity, the participants were split into a “hypermobile” group (4/9 or more on the Beighton Scale assessment) and a “control” group (<4 points on the Beighton Scale). All data were normally distributed (P > 0.05). For statistical purposes, a clearer indication of the magnitude of matching errors was obtained, making all values positive, giving the modular error that had magnitude but not direction.
The study was approved by the Ethics Committee of University College, London, and all participants signed informed consent for all aspects of the research before starting. All data were stored under the provisions of the Data Protection Act 1998.
The sample comprised 20 flautists (4 male and 16 female subjects). Eight (n = 8) were enrolled on a bachelor of music degree at the London music institution mentioned before, 4 were professional flute players, and the remaining 8 were semiprofessionals or playing the flute as a hobby after some years of study. Participants were aged between 20 and 49 years, and the majority were right-handed (only 3 were left-handed or had ambidexterity).
Aspects of the History Determined by Questionnaire
Half of the 20 participants had experienced some injury related to playing the flute or pain severe enough to seek medical advice. Several players reported more than 1 performance-related musculoskeletal problem. These ranged from overuse-type muscular pain to tendonitis and joint pain. One subject previously diagnosed with hypermobility had needed to temporarily reduce or suspend practice because of pain or discomfort. Nearly all players attributed their problems to the postures adopted while playing the flute and long hours of practice.
Flute performance requires the instrumentalists to perform with their arms elevated and abducted and to adopt an asymmetric posture, placing the arms and the instrument to their right. This lateral position can be a source of muscular tension and contractures, as the center of gravity is transferred to the right (Figs. 5 and 6).
The common sites of pain found in this study were the left shoulder, the neck, the left hand (especially the wrist and the index finger; Fig. 7) and the right little finger (Figs. 8 and 9).
Also, all participants who had pain were students or were playing irregularly as a hobby; none of the professionals had pain. This finding might suggest that long hours of irregular practice can cause pain. Regularity may therefore be beneficial, promoting adaptation and endurance, but needing some recovery periods as well to allow the body to adapt. Irregularity may be dangerous, as well as bad scheduling and suboptimal technique.
Seven of the participants had noticed some joint laxity before starting to play the flute, and all noticed an increase in joint flexibility of their fingers, which they attributed to their flute practice. All of these had suffered pain and also related it to their poor performance.
As with all instruments, there is a certain amount of individual variation as to weight bearing, support, and finger options. Although instruments are of uniform size, performers vary in size and shape. Although there are obvious essential requirements ergonomically when playing the flute, the researchers became aware that anatomical differences among the sample had an influence on the site and frequency of pain.
Components of Beighton Scale
Of the 20 participants who completed the testing, only 8 participants (7 female and 1 male subjects aged between 21 and 46 years and with a mean age of 28 [SD, 9] years) had a Beighton Scale score greater than or equal to 4/9. Overall, Beighton Scale scores ranged from 4 to the maximum 9 points. The rest of participants (9 female and 3 male subjects) were used as control subjects, age matching them as closely as possible to the hypermobile subjects. Control subjects scored less than 4 points on the Beighton Scale, with a range 0 to 3 (mean, 1.67).
In addition, the particular maneuvers that they were able to perform were also documented. Seven of 20 subjects were able to passively appose at least 1 thumb to the forearm (4 of them could do both). Fourteen of 20 subjects were able to passively dorsiflex at least 1 fifth finger at the MCP joint to 90 degrees or more, generally the nondominant left (8 of them could do both). Eight of 20 subjects also were able to hyperextend at least 1 (the left) knee by 10 degrees or more (6 of them could do both). Eight of 20 subjects were able to hyperextend at least 1 (the left) elbow by 10 degrees or more (4 could do both). Forward flexion of the trunk so that the palms of the hands rested flat on the floor was achieved by the smallest proportion of subjects (4 of 20).
Passive Hyperextension of the Fifth-Finger DIP, Thumb IP, and Index-Finger MCP
Eleven participants showed greater extension of the right DIP joint of the little finger (mean rank, 9.86), against 7 that showed greater extension on the left side (mean rank, 8.93). Two participants did not report differences between hands. Those findings were not statistically significant (Z = −1.005, P = 0.315). Eight participants showed greater extension of the right IP joint of the thumb (mean rank, 8.31), against 10 who showed greater extension on the left (mean rank, 10.45). Two participants did not report differences between hands. Those findings were not statistically significant (Z = −0.830, P = 0.407). Eight participants showed greater extension of the right MCP joint of the index (mean rank, 8.56), against 11 who showed greater extension on the left (mean rank, 11.05). Only 1 participant did not report differences between hands. Those findings were not statistically significant (Z = −1.069, P = 0.285). Thus, no statistically significant differences were found (P < 0.05) between hands for any of the joints evaluated.
There was a statistically significant difference for the amount of hyperextension of the DIP of the fifth finger on the left hand (H2 = 5.361, P = 0.021) between hypermobile and nonhypermobile participants, with a mean rank of 14.25 for the hypermobile and 8 for the nonhypermobile. There were no statistically significant differences when performing the test for gender and handedness (P > 0.05).
Comparison of the modular error for the hypermobile and control groups was performed (Table 2). About 15% of the errors made by the hypermobile group were greater than 15 degrees in magnitude (12 of 80 matches), compared with the control group, where slightly over 17.5% of the errors were greater than 15 degrees in magnitude (14 of 80 matches). The mean modular error for the hypermobile group was 159.25 (SD, 44.219), which was greater than that of the control group, 155.13 (SD, 68.250), which might infer that the hypermobile group had less accurate proprioception than the control group, but these findings did not reach statistical significance (P = 1).
At different angles (30, 50, 25, −10, −40, 65, −30, 10, and −10 degrees), the errors made by the hypermobile and control groups were not significantly different either.
Hypermobile Flautists Versus Control Subjects
Range of Movement
One study looking at normal passive ranges of motion14 in a normal population (nonhypermobile) found that the MCP joint of the index had an average amount of hyperextension of 14 to 29 degrees for male subjects and 33 to 56 degrees for female subjects. In our study, this average amount increased to 63 to 82 degrees for male subjects and 69 to 74 degrees for female subjects.
The same study found that the DIP joint of the little finger had an average amount of hyperextension of 7 to 15 degrees for male subjects and 14 to 21 degrees for female subjects. In our study, this average amount increased to 32 to 35 degrees for male subjects and 35 to 36 degrees for female subjects.
That study did not evaluate the range of passive motion of the IP joint of the thumb, but taking those 2 examples, one could suggest that flute players have more mobility than do the general population in the finger joints particularly stressed when playing the instrument.
Of the 20 participants who completed the testing in our study, only 8 had a Beighton Scale score greater than or equal to 4/9, which could suggest that flute players do not have generalized hypermobility.
Early studies comparing proprioceptive acuity between hypermobile and control groups10 found that 0.41% of matching errors made were greater than 15 degrees in magnitude for the hypermobile, against 0.21% of errors made by control subjects. Both groups of flute players studied in this project had matching errors more similar to the control subjects than the hypermobile group (15% for hypermobile and 17.5% for control subjects), suggesting that flute players, even if hypermobile, have very accurate proprioception.
The Beighton Scale gives an indication of hypermobility in certain joints (joint laxity) and is frequently used as a quick test in clinic but is of very limited value in measuring flexible fingers because only a few joints are tested. For more accurate measurements of joint laxity in the fingers of instrumentalists, therefore, more sensitive tests are needed, hence the use of the hyperextensometer and the goniometer.
The qualitative data from this study have suggested a number of possible predictors of finger injuries in flute players:
(1) Long hours of practice without recovery periods may be dangerous and can cause pain. Professional flute players practice for fewer hours than do students, which might suggest that it is not the actual amount of study but its quality that is key to achieve excellence.
(2) Irregularity can be detrimental for the flautist’s health as well. All the amateur flautists playing as a hobby had some sort of discomfort when playing. Regularity is crucial, as well as good planning and frequent reviews of the technique.
The health-promoting behavior of music performers is of particular interest, given the physical and emotional demands of expert music making, such as the constant demand for perfection, long periods of intense practice in uncomfortable postures, high competence, financial insecurity, and high levels of anxiety, all of which can be risk factors and may lead to the development of medical problems. A recent study has found that even though they acknowledge this, musicians tend to score poorly in aspects such as health lifestyle (health responsibility, physical activity, and stress management) in relation to other nonmusician populations.15
(3) The asymmetric flute position. The common sites of pain found in this study were the left shoulder, the neck, the left hand (especially the wrist and the index finger), and the right little finger. These findings are in line with other studies done in this field where hand and arm symptoms were common among professional flautists and students,6–8 and shoulder and wrist pain rates were higher in instrumentalists requiring greater arm elevation as occurs with the flute.16
(4) Hypermobility. All the participants with a 4/9 Beighton Scale score or more were complaining of discomfort and muscular/joint pain while playing.
Our main hypothesis that flautists’ finger joints were more mobile than those of the general population was confirmed. Like other groups of musicians who have been assessed for risk injury with hypermobility, the group of flute players had big ranges of movement in the fingers and hand, probably because of the long hours of practice and repetitive movements, as well as the sustained stress in the joints evaluated, which take the weight of the instrument. The hypothesis that flautists have hypermobile finger joints but do not have generalized hypermobility was also confirmed. More than the half of the participants could not perform more than 3 maneuvers of the Beighton Scale. The hypothesis that flautists, even with very mobile small finger joints, have very accurate proprioception (greater than those of control subjects) was also confirmed. It is clear therefore that the amount of practice and the years of training enhance proprioception as well as mobility.
What is interesting is that proprioception does not always equate with body awareness, which needs to become a key aspect in the training of flautists. Teachers must allow time for this in their classes to improve body awareness and technique, as it is a strong predictor of good posture, body efficiency, and musical excellence and can be the best way to prevent discomfort and injuries from the elementary levels of performance.
This study emphasizes the need for future studies in the area of the finger joint laxity, perhaps among different instrumentalists. Further research, both qualitative and quantitative, is required both in the prevention of hand and arm injuries and into the types of interventions and approaches to manage these, in training and in planning musicians’ lifestyle in general. Joint laxity is a phenomenon deserving further attention, particularly given the high prevalence of musculoskeletal and repetitive overuse-related pain in hypermobile musicians. Musicians provide an ideal model for the study of the relationship between joint flexibility and proprioceptive capacity, both inherited and/or acquired through regular practice. String players, where the demands placed on the 2 arms are quite different when compared with wind and keyboard players, are likely to be especially informative in that they act as their own control subjects.
Ultimately, professional performing artists, especially instrumentalists, are athletes of the hands and arms, although this complex area, perhaps the pinnacle of occupational rheumatology, has never attracted as much research interest or investment as the occupational ills of sportspeople.
© 2014 by Lippincott Williams & Wilkins, Inc.