Clinical Journal of Sport Medicine, 12:107–112
© 2002 Lippincott Williams & Wilkins, Inc., Philadelphia
Comparing Injuries of Spin Bowling with Fast Bowling in
Young Cricketers
Peter L. Gregory, MB BS, Mark E. Batt, MB Bchir, and W. Angus Wallace, MB ChB
Center for Sports Medicine, Queen’s Medical Center, Nottingham, England
Objective: To compare the incidence and anatomic distri-
bution of injuries sustained in spin and fast bowling in young
cricketers.
Design: A prospective cohort study. Physical characteristics
and retrospective data (sporting involvement and previous in-
juries) from young cricketers were recorded. The cohort of
young cricketers was divided according to style of bowling into
a group of spin bowlers and a group of fast bowlers. Data from
the spin bowler group was compared with the data from the fast
bowler group to assess whether these groups were matched. A
prospective study of injuries sustained by the bowlers was then
undertaken. Data regarding cricket played and injuries sus-
tained were collected by telephone questionnaire every 6 weeks
for 6 months from each bowler.
Setting: Bowlers were recruited from young cricketers train-
ing at Centers of Excellence of 3 “First Class” Counties in
England in January 1998.
Participants: There were 42 spin bowlers and 70 fast
bowlers. The mean age was 14.9 years (range 9 to 21 years).
Main Outcome Measures: Injuries caused by bowling and
interfering with bowling are included in incidence data. The
number of deliveries bowled in matches and practice is used as
the denominator for the reported incidence.
INTRODUCTION
A cohort of young cricketers was studied to discover
the incidence and anatomic pattern of spin bowling in-
jury. The findings were compared with data from fast
bowlers of the same squads. The study was undertaken to
address the lack of data in literature on spin bowling
injuries.
In cricket, bowlers can be categorized as fast bowlers
or spin bowlers. All bowlers propel a 5.5 oz ball towards
a batsman or his wickets, but a spin bowler imparts ro-
tation to the cricket ball, which will make the ball deviate
from its original direction of flight when it hits the
ground. The spin bowler achieves this by rapid flexion of
the fingers around one side of the ball, called “finger
spin,” or by rapid movements of the wrist as he or she
Accepted December 3, 2001.
Address correspondence and reprint requests to Peter L. Gregory,
MB BS, MSc Sports Medicine Center for Sports Medicine, Orthopae-
dic & Accident Surgery, C Floor West Block, Queen’s Medical Center,
University Hospital, Nottingham NG7 2UH, England.
107
Results: Telephone follow-up was achieved when planned
on 98.2% of occasions. There were 29 injuries meeting the
criteria above. The incidence of injury in spin bowlers was
0.066 per 1,000 balls and 0.165 per 1,000 balls in fast bowlers
(p ⳱ 0.097 Wilcoxon rank sum test). The incidence of injury
(per 1,000 balls) at various anatomic sites in fast bowling was
knee 0.057, ankle 0.043, low back 0.029, and shoulder 0.007.
In spin bowlers, the site incidence was shoulder 0.055 and low
back 0.011. The percentage with injuries at ankle, knee, and
shoulder was significantly different (95% confidence intervals)
for fast and spin bowlers. A significant difference was not
found for lower back injuries.
Conclusions: Incidence of injuries in fast bowling is greater
than in spin bowling (but this was not a significant difference
within the limits of this study). Low back injuries in fast
bowlers have been the subject of published research. However,
injuries in spin bowling have not previously been described,
and this study indicates that shoulder injuries in wrist spinners
merit further study.
Key Words: Sport—Cricket—Bowling—Injury—
Incidence.
Clin J Sport Med 2002;12:107–112.
releases the ball, called “wrist spin.” Detail of the bio-
mechanics involved in these bowling actions is not re-
corded in the literature. Spin bowling contrasts with fast
bowling when bowlers try to beat the batsmen by com-
binations of speed of delivery and deviations in flight,
usually brought about by keeping the stitched seam of
the cricket ball vertical and pointing towards the bats-
man. Deviations may occur before the ball bounces,
called “swing” or after, called “cut” (Figure 1).
For 30 years, spin bowling, particularly wrist spin, was
unfashionable. It was thought to be ineffective in limited
over cricket introduced in the 1960s, and young cricket-
ers were inclined to model themselves on successful in-
ternational fast bowlers. More recent success at the in-
ternational level by wrist spin bowlers has renewed in-
terest in spin bowling. Harris1 studied 99 South African
cricketers (mean age 20.2 years) and found the preva-
lence of low back pain to be 48.1% in 27 spin bowlers,
compared with 75.6% in 37 fast bowlers. There are no
other published studies describing the injury pattern as-
sociated with spin bowlers, and so it is difficult to predict
108 P. L. GREGORY ET AL.
FIG. 1. Diagram showing fundamental difference between
method of spin bowling and fast bowling with likely difference in
deviation after bounce.
the influence this increase in spin bowling will have on
our sports injury clinics.
Much more is already known on fast bowling, where
technique problems and overuse have been implicated in
the etiology of stress fractures of the pars interarticu-
laris.2–4 Back injuries of fast bowlers have been studied
in the 1980s and 1990s. Foster2 found that bowlers who
bowled more than the mean number of overs in a study
of 82 young fast bowlers had a higher prevalence (58%)
of stress fracture and soft tissue back injury than the
other bowlers (38%). Elliott4 found that bowlers with
spinal abnormalities were significantly older that those
without. Elliott did not report the amount of bowling, but
postulated “a greater tendency to display abnormal ra-
diologic features with more years of bowling.” The En-
gland and Wales Cricket Board (ECB) responded in
1992 by drafting directives5 that restricted the number of
balls bowled by young fast bowlers in matches and train-
Clin J Sport Med, Vol. 12, No. 2, 2002
ing. The matches have not reduced in length, and so an
impact of the directives is that young spin bowlers will
bowl more. There is no spin bowling injury incidence
data in the literature, and no studies relate amount of
bowling to injury in spin bowlers.
METHODS
Coaches at three English County Cricket Club Centers
of Excellence directed us to 115 young cricket bowlers
who were asked if they were willing to take part in this
study. Two cricketers refused, believing it would take up
too much of their time. Thus, 113 bowlers were recruited
in January 1998. These were divided into 70 fast bowlers
and 42 spin bowlers by their own description and con-
firmed by coach analysis. One bowled a mixture of spin
and fast and was excluded from the comparative analy-
sis. Other cricketers training at these Centers of Excel-
lence were excluded, as they did not bowl. The 42 spin
bowlers were made up of 27 wrist spin bowlers and 15
finger spin bowlers. The techniques were different, and
so injuries in these subgroups were compared where in-
cidence of injury to a particular anatomic site was higher
in the spin bowling compared with the fast bowling.
All the bowlers underwent a structured interview and
physical. They were asked about involvement with other
sports, physical training, cricket played last year, and
previous injuries. This retrospective data and measure-
ments of height and body mass were used to assess how
well the groups (fast bowlers and spin bowlers) were
matched.
All bowlers were followed-up by telephone question-
naire every 6 weeks for 6 months. This included 3
months of preseason (indoor) training and the first 3
months of the 1998 cricket season. This provided the
prospective data and consisted of details of the cricket
played and injuries sustained. Centers of Excellence pro-
vide cricketers with a logbook, and the details recorded
on bowling and injuries provided a useful aid in recall in
this study. In cricket, six consecutive balls bowled by
one bowler are called an “over.” Bowlers stated how
many overs and extra balls they bowled in each match.
They provided an estimate of the time bowling in the
nets and the number of other bowlers sharing a net with
them. The number of balls bowled in 18 5-minute peri-
ods of observed net sessions averaged 22 balls. This
factor was then used to calculate the number of balls
bowled by individual bowlers during net sessions. From
this information the total number of balls bowled in
matches and practices (nets) was calculated.
The bowlers were asked to report every injury and
detail the circumstances. This included whether the in-
jury occurred while batting, fielding, or bowling. Some-
times there was no moment of obvious trauma, but the
bowler was still convinced that bowling had led to the
injury. Only bowling injuries were recorded in incidence
data in this study. Injuries were self-reported during the
telephone interviews and did not necessitate verification
by a health professional for inclusion. Injuries were
graded according to the classification in Table 1. Report-
 COMPARING INJURIES OF SPIN VS. FAST BOWLING 109

TABLE 1. Classification of injury severity the amount of cricket played or for the amount of batting,
bowling, and fielding done in these matches and prac-
Grade I Pain following bowling
Grade II Pain during bowling tices. However, the spin bowlers spent 555 overs fielding
Grade III Pain impairing bowling performance in the close-catching positions, while fast bowler spent
Grade IV Pain preventing bowling only 236 overs in these positions; this difference was
statistically significant (p⳱ 0.01).
Previous Injuries
ing of minor injuries was likely to vary, so injuries not
Past injuries, particularly with on-going problems,
severe enough to impair bowling performance or to pre-
might lead to bowlers being more susceptible to injury in
vent bowling were excluded from the analysis.
the study period. The bowlers were asked to report all the
Microsoft Excel (Redland, WA, U.S.A.) was used to
injuries they could remember. The total number reported
store the data, and statistical significance was sought at
was 221. Spin bowlers reported 1.62 injuries per bowler,
the 5% level using Wilcoxon’s test on SPSS (SPSS, Inc.,
while fast bowlers reported 2.13 injuries per bowler. Ap-
Chicago, IL, U.S.A.). Height and body mass appeared to
proximately 28.6% of the fast bowlers and 23.8% of the
be distributed normally and were compared with a two-
spin bowlers had an ongoing injury at the time of the
sample t test.
structured interview (difference 4.8–95% confidence in-
terval (CI) for the difference +/−16.9%).
RESULTS Forty-eight percent of the 221 reported injuries pre-
Physical Characteristics vented the bowlers bowling at some stage. Fast bowlers
The mean age of the 113 bowlers was 14.9 years reported 1.16 injuries preventing bowling per bowler
(range 9.5 to 21.5 years). The mean age of the fast compared with 0.60 injuries per spin bowler (p ⳱
bowlers was 1 year greater than the mean age of the spin 0.015). Differences in proportions of bowlers reporting
bowlers (p ⳱ 0.035). Characteristics of this cohort are less severe injuries were not statistically significant. Low
detailed in Table 2. back injuries had been more common in fast bowlers
(0.40 per bowler) than spin bowlers (0.12 per bowler) (p
Retrospective Data ⳱ 0.02). Differences at other anatomic sites were not
Retrospective data is presented to compare the influ- statistically significant.
ence of possible confounding variables in the fast and
spin bowler groups. Prospective Study
Other Sport and Training in Previous Year There were 444 telephone contacts with the bowlers
The bowlers had spent a mean of 22.4 hours compet- during the study. No bowler was contacted less than
ing in other sports in the preceding 12 months. They had three times. The mean time between contacts was 43.3
been involved in 43.4 practice sessions, not including days. The bowlers were followed-up for 174.2 days
physical education lessons at school. In addition the (mean). The bowlers played a mean of 15.8 matches
bowlers spent 34.5 hours exercising to improve their fit- during the study period, which amounted to 17.7 days of
ness to play cricket. Differences between the activity of cricket or 1,338 overs of cricket. The mean number of
spin and fast bowlers were not found to be a statistically balls bowled by each bowler during the study period was
significant, except for the amount of time spent on re- 2,046 (range 571 to 8,045). Table 3 shows where this
sistance training in multigyms. The faster bowlers aver- bowling occurred. The fast bowlers bowled a mean of
aged 19.3 hours in the year on this aspect of their train- 1,995 balls, and spin bowlers bowled a mean of 2,163
ing, the spin bowlers only 2.1 hours (p < 0.01). balls (p ⳱ 0.086).
Cricket Involvement During Previous Season Injuries
The bowlers participated in 47 outdoor cricket Incidence. During this study, 95 cricketing injuries
matches in 1997 (median). They had attended 20 outdoor were reported. Of these injuries, 44 were attributed to
and 10 indoor practices. No statistically significant dif- bowling. Of the injuries that were attributed to bowling,
ferences were found between spin and fast bowlers for 29 interfered with bowling (Grade III or IV) and are
represented in the following data. The period prevalence
of Grade III or IV bowling injuries to fast bowlers during
TABLE 2. Physical characteristics of fast bowlers (n = 70) the study was 0.338, and to spin bowlers, 0.146 (p ⳱
and spin bowlers (n = 42) 0.122). The incidence of these injuries to fast and spin
Mean (SD)
Characteristic
type of bowler All Fast Spin Probability
TABLE 3. Balls bowled per bowler (n = 113) in the study
Age (years) 14.9 (2.5) 15.3 (2.4) 14.3 (2.5) 0.035
(January to July 1998)
Height (meters) 1.68 (0.14) 1.71 (0.14) 1.64 (0.14) 0.021
Weight (kg) 60.4 (15.2) 62.6 (14.5) 57.2 (15.9) 0.072 Outdoor Outdoor Indoor Indoor
BMI 21.0 (3.0) 21.1 (2.5) 20.8 (3.7) 0.521 matches practices matches practices
Probability calculated from Student’s t-test. Balls bowled (mean) 604 561 22 859
BMI, body mass index.

Clin J Sport Med, Vol. 12, No. 2, 2002

110 P. L. GREGORY ET AL.
bowlers was 0.162 and 0.068 per 1,000 balls bowled,
respectively (p ⳱ 0.096).
Site of Injury. The four most common sites of injury
reported in this study period were ankle (17), knee (15),
shoulder (14), and lower back (12). The incidence of
Grade III or IV fast bowling injury (per 1,000 balls) at
various anatomic sites was knee 0.057, ankle 0.043, low
back 0.029, and shoulder 0.007. In spin bowlers, the site
incidence was shoulder, 0.055, and low back, 0.011.
Table 4 gives the proportion of bowlers sustaining Grade
III or IV bowling injuries at these four most commonly
affected sites. The 95% CI for the standard error of the
difference in percentage of fast and spin bowlers indicate
a statistically significant difference at the ankle, knee,
and shoulder, but not the lower back. The ankle and knee
were injured more by fast bowling, while the shoulder
was injured more by spin bowling.
Five of the six Grade III or IV fast bowling ankle
injuries were sprains, while all eight knee injuries of this
severity were attributed to fast bowling begun without
obvious trauma (five undiagnosed anterior knee pain,
two Osgood-Schlatter disease, and one patellar tendon-
itis). Approximately 10.0% of fast bowlers and 11.9% of
spin bowlers reported low back pain (all grades) during
the study period. However, Grade IV low back pain oc-
curred only in fast bowlers (three subjects). One 13-year-
old had a pars interarticularis stress fracture confirmed
on isotope bone scan and computed tomography scan.
Ten percent of fast bowlers and 16.7% of spin bowlers
developed shoulder injuries (all grades) during study. No
fast bowler ascribed their shoulder injury to bowling,
while all five spin bowlers who developed Grade III or
IV shoulder injuries attributed their problem to bowling.
Four fast bowlers blamed throwing the cricket ball
(while fielding) for their shoulder injury and said their
bowling was unaffected. There were three Grade IV
shoulder injuries in the spin bowlers. These most severe
injuries occurred in wrist spinners.
DISCUSSION
Follow-up by telephone questionnaire proved to be a
satisfactory alternative to postal questionnaire. Approxi-
mately 98.2% of planned contacts were achieved. Re-
peated follow-up at 6-week intervals was preferred to the
single postal questionnaire at the end of the study period
used in the cricket studies of Stretch6,7 and Harris,1 as the
bowlers had less time to forget the required details.
The strength of a comparative prospective study de-
TABLE 4. Proportion of bowlers sustaining injuriesa at four anatomical sitesb
Type of
bowler Ankle
Fast 8.6%
Spin 0.0%
% Difference 8.6%
95% confidence intervals for % difference 2.0–15.1%
a
Grade III or IV injuries caused by bowling.
b
The four most commonly injured anatomical sites caused by bowling are detailed.
Clin J Sport Med, Vol. 12, No. 2, 2002
pends on groups being well matched for all confounding
variables. Statistically significant differences between
the fast and spin bowling group were found in 14 of 111
variables tested. Chance would predict about six vari-
ables being significantly different, as statistical signifi-
cance was sought at the 5% level. The difference be-
tween the mean age of the fast bowlers and spin bowlers
was 1 year. We expected taller cricketers in the fast
bowler group as this gives a physical advantage, but in
fact the mean height and body mass for each group was
on the 50th centile for their respective age. Being older,
and consequently taller and heavier, might lead to fast
bowlers having a higher injury rate. Spending more time
fielding in close-catching positions would make the spin
bowlers less susceptible to injuries to the throwing shoul-
der. In several ways, the fast bowlers had the worst past
injury profile. This might have led to a greater incidence
of injury in the study period, with recurrences or greater
susceptibility to new injury. Overall, we predict that the
confounding factors could exaggerate the injury rate in
the fast bowlers relative to the spin bowlers in this study.
Incidence
Information on occurrence of injuries is more useful if
exposure to injury risk is quantified.8 Hours of matches
were recorded, which allowed calculation of incidence
data in a comparable way to published studies of injury
in other sports, where incidence of injury is often quoted
per 1,000 hours of playing time. This is useful when
comparing different sports for risk of injury. It is also
useful when comparing results from different studies of
the same sport. Pringle9 reported injury rates for three
contact sports in 6-to-15 year olds in New Zealand:
Rugby Union 15.5 per 1,000 player hours, Rugby League
24.5 per 1,000 player hours, and Netball 13.0 per 1,000
player hours. Pringle included any injury causing dis-
comfort after the match and used trained observers at
matches to collect the data. In our study of young crick-
eters, there were 6.8 injuries (any severity) per 1,000
hours of cricket matches. A lower incidence in cricket
might be expected, as “collision” occurs infrequently in
this sport compared with rugby and netball.
The amount of bowling by players in cricket matches
varies considerably, and this is confirmed in both data
from our retrospective and prospective studies. This
study aimed to discover the risks of bowling and not
merely playing cricket, so we have presented incidence
with the denominator as balls bowled in practice and
matches. The incidence of injuries in our study was
Low
Knee back Shoulder
11.4% 5.7% 1.4%
0.0% 2.4% 11.9%
11.4% 3.3% −10.5%
4.0–18.9% −3.8–10.5% −20.7 to −0.3%
 COMPARING INJURIES OF SPIN VS. FAST BOWLING 111

0.125 per 1,000 balls bowled, and the specific rates were while the arm circumducts (see Figure 2). We speculate
calculated for fast bowlers (0.162 per 1,000 balls) and that it may be this internal rotation that may predispose
spin bowlers (0.068 per 1,000 balls). No other published to impingement and injury. The situation may be analo-
study reports this level of detail. gous to freestyle and butterfly swimming. Shoulder in-
Stretch reported a seasonal incidence of injury of jury in wrist spinners merits further research, with kine-
about 50% in schoolboys7 and first league and provincial matic analysis and detailed injury assessment.
cricketers6 in South Africa. Stretch counted injuries oc-
Strengths and Weaknesses of This Study
curring in matches and practices, which prevented
The strengths of this study are its prospective nature
completion of the match or practice session. The sea-
and use of incidence data. The cohort is representative of
sonal incidence of cricketing injuries of this severity ex-
young English bowlers. The 6-month study period is rep-
trapolated from the data collected in our prospective
resentative of a full year, including 3 months training and
study is 48.8 injuries per 100 bowlers. Stretch surveyed
3 months cricket season. Excellent follow-up was
all types of cricketers and found 47.4% of cricket injuries
achieved by telephone questionnaire at frequent intervals
occurred when bowling. In our study, which focuses on
reducing recall problems, and there was no drop-out.
bowlers, the proportion of cricketing injuries (preventing
Minor injuries (Grade I and II) that might be subject to
bowling) caused by bowling was 64.5%.
recall bias were excluded. Detailed assessment and con-
sideration of possible confounding factors were made by
Pattern of Injury
analysis of physical characteristics and retrospective data
We found that fast bowling was more likely to cause
for the two groups. Taken together, the physical charac-
ankle sprains and knee injuries with no obvious trau-
teristics, retrospective data, and prospective data provide
matic cause. Wrist spin bowling was more likely to cause
a detailed description of young English bowlers and their
shoulder injuries. Interestingly, low back pain was
sporting behavior.
equally common in fast and spin bowlers. However,
Weaknesses of the study include failure to eliminate
Grade IV low back injury was associated only with fast
confounding variables. However, we predict these would
bowling.
tend to exaggerate injury rate in fast bowlers, and so
Harris1 did not state the period of time over which he
would not weaken our interpretation regarding injuries in
recorded the prevalence of low back pain. In our study,
young spin bowlers. Though a large number of bowlers
the probability of the bowlers suffering low back pain on
were recruited, injury rates were lower than expected,
any day was 3.5%. Extrapolating data from our study
and so the power of the study was reduced. Injuries were
gives a prevalence of low back pain for the season of
self-reported by the bowlers. They did not undergo ex-
25.7%. Harris studied cricketers who were on average 5
amination by the investigators, and thus a precise diag-
years older than those in our study. Age may affect injury
pattern. Most of the cricketers in the study of Harris
would have reached skeletal maturity, while most in our
study would not. Low back pain is less common in the
general population before skeletal maturity than after.
The bowlers studied by Harris played cricket on hard
South African wickets, which may also have contributed
to the greater stated prevalence of low back pain of
75.6% in fast bowlers and 48.1% in spin bowlers. In our
study, 10% of fast bowlers and 12% of spin bowlers
developed low back pain. Low back pain that prevented
bowling was restricted to fast bowlers.
Stretch7 found 55.5% of bowling injuries were to the
back and 29.5% to the lower limb. We found only 18%
of Grade III and IV bowling injuries were of the trunk,
but 61.5% of the lower limb. Age may have influenced
our high numbers of lower limb injuries, where apoph-
yseal injuries are expected in the skeletally immature.
The two cases of Osgood-Schlatter disease are an ex-
ample of this.
In the last 15 years, research has attempted to identify
the cause of “fast bowlers’ back.” Subsequently preven-
tative measures have been taken, and it could be that as
a consequence, we found a lower proportion of cricketers
with this type of problem in our study than in previously
published research.
The problem of shoulder injury in wrist spin bowling FIG. 2. Drawing of finger spin and wrist spin bowling at point of
has not been reported previously. In wrist spin, the ball release, showing that the bowler’s shoulder in wrist spin is
bowlers appear to rotate the bowling shoulder internally internally rotated.

Clin J Sport Med, Vol. 12, No. 2, 2002

112 P. L. GREGORY ET AL.
nosis for each injury is not offered. However, injuries
were reported by anatomic region. Assessment of bowl-
ing technique was made by coaches only, but did not
involve kinematic analysis, and so has been excluded
from this report.
CONCLUSIONS
The data collected from this cohort of young English
cricketers indicate that injuries that impair bowling and
that are caused by bowling occur more often in fast
bowlers than spin bowlers, though we were unable to
demonstrate statistical significance in this prospective
comparison of 112 cricketers lasting 6 months. Incidence
of low back injuries in fast bowlers was less than in
previous published studies, which have been done in
South Africa and Australia. Low back pain was equally
prevalent in spin bowlers in our study, though the sever-
ity was less. Spin bowling, and in particular wrist spin
bowling, is associated with injuries of the bowling shoul-
der. This association merits further investigation.
Acknowledgment: We thank Head Coaches Stewart Bur-
rows, Andy Hayhurst, and Russell Cobb, and their coaching
teams of Nottinghamshire CCC, Derbyshire CCC, and Leices-
Clin J Sport Med, Vol. 12, No. 2, 2002
tershire CCC Centers of Excellence; Tanya McClean, Depart-
ment of Orthopaedic & Accident Surgery, Nottingham, for sta-
tistical guidance; Linden Cochrane for producing Figures 1 and
2; and 113 young cricketers who conscientiously provided the
data.
REFERENCES
1. Harris I. The prevalence of low back pain in cricketers—an un-
dergraduate epidaemiological study. S A J Physiother 1993;94:
65–66.
2. Foster DH, John D, Elliott B, et al. Back injuries to fast bowlers in
cricket: a prospective study. Br J Sports Med 1989;23:150–154.
3. Elliott BC, Hardcastle PH, Burnett AF, et al. The influence of fast
bowling and physical factors on radiologic features in high perfor-
mance young fast bowlers. Sports Med Train Rehab 1992;3:113–
130.
4. Hardcastle P, Annear P, Foster DH, et al. Spinal abnormalities in
young fast bowlers. J Bone Joint Surg 1992;74B:421–425.
5. National Cricket Coach Initiative. Hitting the Seam. Birmingham:
Warwickshire County Cricket Ground, 1988;4–5.
6. Stretch RA. The incidence and nature of injuries in first league and
provincial cricketers. S Afr Med J 1993;83:339–342.
7. Stretch RA. The seasonal incidence and nature of injuries in
schoolboy cricketers. S Afr Med J 1995;85:182–184.
8. de Loes M. Exposure data—“Why are they needed?” Sports Med
1997;24:172–175.
9. Pringle RG, McNair P, Stanley S. Incidence of sporting injury in
New Zealand youths aged 6–15 years. Br J Sports Med 1998;32:
49–52.