BrJ Sports Med 1996;30:134-139

Low back pain and other overuse injuries in agroup ofJapanese triathletes

Jyri Seppo Olavi Manninen, Mauri Kallinen

AbstractObjective-To document the incidence oflow back pain and other overuse injuriesin a group oftriathletes, and to investigateany associations with various physical andtriathlon related factors.Methods-By means ofa questionnaire, thephysical characteristics, training habits,and the incidences of overuse injuries of 92Japanese triathletes (70 males, 22 females)were documented. Student's t and x2 testswere used to determine the significance ofany associations with injury incidence, aswell as differences between subjects experi-encing or not experiencing low back pain inthe previous year.Results-Low back pain was experiencedby 32% ofsubjects in the previous year. Themajority (54%) of low back pain episodeslasted under seven days, suggesting mainlysoft tissue involvement, and 19% lastedover three months, suggesting involvementof the intervertebral discs. Weekly trunkflexor muscle training frequency was signi-ficantly greater (P = 0.035) for the low backpain subjects. Close to significant differ-ences for average weekly cycling time,trunk flexor muscle training time, and lowintensity aerobic rining, as well nearsignificant associations for weight trainingand average weekly triathlon taining load,were also found. No other factors weresignificantly associated with low back pain.Low back injuries accounted for 28% of allinjuries. Only the knee was a morecommon single site ofinjury (33%/o).Conclusions-The three most commoninjuries suffered by the triathletes were ofthe knee, back and shoulder. The low backpain suffered by many triathletes could beof a potentially serious nature. It issuggested that cycling is a major riskfactor for low back pain in triathletes.(BrJ Sports Med 1996;30:134-139)

Department ofHealthSciences, University of'JyvdskylA, FinlandJ S 0 ManninenM KallinenThe study was carried outat the Institute of SportsMedicine and Science,Agui, JapanCorrespondence to:Jyri Seppo Manninen,1046 NW 9th Street,Corvallis Oregon 97330,USA

Accepted for publication8June 1995

Key terms: low back pain; triathlon; injury; cycling

Triathlon is an endurance sport involvingswimming, cycling, and running. It has of-ficially existed only for about 20 years, andduring this time has developed very rapidlywith regard to training strategies and cyclingtechnology. Because of the short history of thesport, it has been very difficult to evaluate, or

even predict, its long term physical effects (thatis, sports related factors) on the participants. Itis really only in the new or younger generationof triathletes, for many of whom triathlon was

their first sport, that these long term effects canbe monitored and investigated more accu-rately. Data on triathlete injuries are sparse,and the most recent studies were performedsome five years ago'`7 This, along with theabove mentioned rapid development of thesport - especially in recent years - indicates thestrong need for further more comprehensiveinvestigation into triathlon related injuries.

It has been argued that triathlon, as amultisport event, causes less overuse injuriesthan single sports, because of the more evendistribution of loads over the body's musculo-skeletal structures. Despite this, triathletes stillsuffer from a high degree of overuse injuries.These overuse injuries also account for asmuch as 85% of all injuries suffered.3' Onecommonly injured site is the low back. 15The main purposes of this study were to

document the incidence of low back pain andother overuse injuries in a group of triathletes,and to investigate the associations betweencertain physical and triathlon related factorsand the incidence of low back pain.

MethodsA pilot study was conducted to design atriathlon injury questionnaire to obtain data onthe physical characteristics oftriathletes (that is,age, height, weight, body mass index), theirtraining habits (that is, weekly training timesand distances, type of training performed andat what intensities, etc), as well as on theirtriathlon related injuries (their incidences andpossible causes), in particular low back pain. Inaddition, test-retest reliability of the question-naire was determined. It was found to be highlyrepeatable (Manninen J: Triathlon and lowerback pain: a pilot study. Unpublished paper forphysiotherapy masters programme. UniversityofJyviskyla, Finland, 1994). The questionnairewas then employed in this present study forobtaining the same type of data from theJapanese triathlete subjects.

Representatives of various triathlon clubsaround Japan were contacted about the presentstudy, and were told that it involved triathlontraining methods and injuries. Once informedconsent had been obtained from the represen-tatives, questionnaires were mailed out. Thenumbers allotted to each club were based onestimates of club member numbers. The clubrepresentatives were asked to distribute thequestionnaires to those club members willingto take part, and to then return both completedas well as unused questionnaires.During March 1994, 185 questionnaires

were mailed out to the representatives of 10

134

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from

Low back pain in Japanese triathletes

31

8'M~~Female

8BO Male

23

8

0 42

31-35 3640 41-45 48-50 ,51

Age group (years)The age distribution of the Japanese triathlete subjects.

selected triathlon clubs. Most were located inthe Kinki, Chubu, and Kanto regions of Japan.Of these questionnaires, 109 were returnedcompleted. The questionnaires returned by 17subjects were excluded on the basis that theydid not indicate that they were training all thethree sports. Therefore, by definition, theywere not triathletes. Due to the distributionmethod employed, the completion rate ratherthan actual response rate was calculated. Thetrue completion rate for the triathlete subjectswas 550/o (92/168).

Statistical analyses included descriptivemethods, Student's t tests, and x2. The level ofstatistical significance for both the Student's ttest and x2 analyses was set at P = 0.05. TheExcel 4 0 and SPSS 6-0 computer softwarepackages were employed to carry out theseanalyses.

Table 1 Physical characteristics of triathlete subjects. Values are mean (SD), [range]

Male (n = 70) Femnale (n = 22) Total (n = 92) P

Age (years) 31-6 (7 4) 30-1 (7 3) 31-3 (7 4) 0-39[19 to 52] [20 to 56] [19 to 56]

Height (m) 170-7 (65) 158-0 (52) 167-7 (8-2) < 0 01[154 to 185] [143 to 165] [143 to 185]

Weight (kg) 63-8 (7 5) 51-4 (4-6) 61-0 (8 7) < 0 01[46-0 to 85 0] [40 0 to 64-0] [40 0 to 85 0]

Body mass index (kgm2) 218 (0 2) 20-8 (0 3) 21-6 (0 2) < 0 01[17-3 to 30-1] [19-0 to 23 9] [17-3 to 30-1]

Table 2 Average weekly training times and distances of triathletes. Values are means(SD), [range]

Male (n = 70) Female (n = 22) Total (h = 92) P

Times (hours)Swim 3-6 (2 0) 3 9 (1 9) 3-7 (1-9) 0-52

[0 3 to 10-] [10 to 90] [0 3 to 10-0]Bike 5-0 (2 7) 4-8 (2 5) 5 0 (2 6) 0 73

[0-5 to 12-0] [1-0 to 100-] [05 to 12-0]Run 4-3 (2 0) 4-6 (2 3) 4-3 (2 0) 0-58

[0-3 to 90] [1-0 to 13-0] [0-3 to 13-0]Weights 1-4 (1-4) 0-6 (1 1) 1-2 (1-3) 0-012

[O to 50] [0 to 4 0] [O to 50] [P < 0 05]Circuits 005 (1 0) 0-5 (0 8) 0 5 (10) 0-98

[O to 5-0] [O to 2 5] [O to 50]Total 14-7 (64) 14-4 (5-1) 14-6 (61) 0-87

[2-2 to 30 0] [5-0 to 24 5] [2-2 to 30-0]Distances (kin)Swim 6-4 (3 7) 6-8 (3 3) 6-5 (3 6) 0-66

[0 3 to 20] [2 to 13] [3 to 20]Bike 123-6 (67 5) 99-8 (52 0) 118-0 (64 7) 0 09

[10 to 300] [30 to 200] [10 to 300]Run 42-0 (24-3) 43-0 (22 5) 42-2 (23 8) 0-87

[1 to 150] [5 to 120] [1 to 150]

ResultsGENERAL

The physical characteristics of the triathletesare described in table 1. Only age did not differsignificantly between male and female subjects.Most (63%) of the triathletes were aged 26 to35 (figure), and most (96%) were in occu-pations involving sedentary or light activities(for example, business, office work, study,research, and medical). Only one subject wasa professional triathlete. Six subjects smokedcigarettes regularly. Three of these subjectssmoked one or less per day, while the otherthree smoked an average of 10 cigarettes perday.The subjects had been training regularly for

triathlon on average for 4-1 years (4 3 formales, 3-3 for females). The average weeklytraining times and distances of the triathletesare described in table 2. There were no signifi-ant gender differences except for weight train-ing, in which females trained significantly less.Seventy three per cent of subjects performedweight training. Only 29% included circuittraining in their overall training programme.Supplementary training (circuit and weighttraining combined), accounted for 12% of totaltraining time. Swimming, biking, and runningaccounted for 29%, 38%, and 33% respect-ively, of total triathlon training time (that is,not including supplementary training).Further, 86% of subjects used triathlon "aero-bars" (special handlebars) on their bicycles.Most of the training was performed at medium(41%) and low intensities (38%), the remain-der being performed at high intensity.5 Overhalf (57%) trained for half and full distanceevents. The remainder (43%) trained forOlympic distance and sprint events. Beforetriathlon training, 27% of subjects participatedin running, 12% in swimming, and only 4% incycling. A regular stretching programme wasfollowed by 94% of subjects. On average, this

a Training intensity levels:Low intensity - between the aerobic and anaerobicthresholds.Medium intensity - at or below the aerobic threshold.High intensity - at or above the anaerobic threshold.

35

30n =92

U)0a7C0

._

*a

Ez

25

201

15

10 [51

135

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from

Manninen, Kallinen

Table 3 Low back pain incidence rates (%) of triathletesubjects

Male (n = 70) Female (n = 22) Total (n = 92)

n % n % n %

Lifetime 45 64 9 41 54 59Yearly 23 33 6 27 29 32Monthly 15 21 3 14 18 20Point 9 13 2 9 11 12

was performed 6&8 times per week for 13-4minutes each time.

LOW BACK PAINThirty two per cent of subjects had experi-enced low back pain in the previous year. Theincidence rates are given in table 3. There wereno significant differences between males andfemales, although the males had a notablyhigher life time incidence rate (P = 0052). Thegreatest incidence of low back pain occurredduring the mid-season training phase (44%). Ahigh proportion (23%) also occurred duringthe competition season. The average age thatlow back pain was first experienced was 23-7years (23-8 for males, 23-0 for females). Initiallow back pain episodes began for no particularreason, either suddenly or little by little, in 51%of subjects. Twenty two per cent stated thattheir initial pain episode had begun suddenly,but was due to some definite incident. Themajority of pain episodes (54%) lasted underseven days. Twenty seven per cent lasted oneweek to three months, with the remainder(19%) lasting over three months. Pain hadmainly been felt in the lower and upper back(58%), but 32% had, in addition to their lowback pain, experienced lower limb symptomsin either one or both legs. The remainder hadexperienced lower limb symptoms in theabsence of low back pain. Because of their lowback pain, over half (56%) had stopped workor training for some time during the previousyear. Four per cent had been unable to trainat all for extended periods of time. Low backpain symptoms decreased in 45% of affectedsubjects over their period of triathlon training.Twenty nine per cent had not noted any overallchanges.Average weekly trunk flexor muscle training

frequency was significantly greater in thosesubjects who had experienced low back pain inthe previous year (P = 0035). Other resultsalso indicated close to significant differences (ttests) between the low back pain and non-low-back-pain subjects, or associations (X2 tests)with the incidence of low back pain in theprevious 12 month period. These were: averageweekly trunk flexor muscle training time(P = 0 055), average weekly cycling time(P = 0 065), proportion oflow intensity aerobictraining (P = 0-09) (t tests); performance ofweight training (P = 0054), and average weeklytriathlon training loada (P = 0095) (X2 tests).aThe differences between low back pain and

Table 4 Injury rates of triathlete subjectsNumber of Male (n = 70) Female (n = 22) Total (n = 92)injuries

n % n % n %

0 18 26 8 36 26 281 29 41 8 36 37 402 15 21 5 23 20 22>=3 8 11 1 5 9 10

symptom-free subjects for gender, age, jobactivity level, flexibility, other training methods,and prior participation in other sports were notsignificant. Also, neither biking position, interms of the degree of trunk flexion in asubject's usual cycling position, nor the use oftriathlon "aerobars" was associated with theincidence of low back pain in the previous year,nor was the presence or number of lower limbinjuries suffered during the previous yearassociated with the incidence of low back painin that year. Seventy four per cent of the lowback pain subjects associated pain with cycling,45% with weight training, and 43% withrunning. The least likely association stated wasfor circuit training (19%).

OTHER INJURIESSeventy two per cent of subjects had sustainedat least one musculoskeletal injury related totriathlon in the previous year. Of this group,30% had sustained two injuries, and 14% hadsustained three or more injuries (see table 4).Most injuries involved the lower limb (61%),of which the knee was the most common siteof injury (54% of all lower limb injuries).Running was the most common activity statedas a cause oflower limb injury. Specific reasonsincluded overtraining, faulty running tech-nique, a hard running surface, and inadequatestretching. Only 8% of injuries were of theupper limb, all involving the shoulder. Swim-ming related factors were stated as the causesof all shoulder injuries. They included poorswimming technique, inadequate stretching,and inadequate warm up. An injury involvingthe spinal region had been sustained by 32%of subjects (lower back injuries 28%, neck/cervical region 4%). Only the knee was a morecommon single site of injury (33%). Table 5classifies the injuries according to anatomicalregion.

DiscussionThe subjects in this study comprised a totallyhomogeneous group by race, that is, they were

Table 5 Anatomical distribution (%0) of injuries (n = 105)in triathlete subjects (n = 66)

Anatomical site ofinjury Frequency

n %

Knee 35 33-3Low back 29 27-6Shoulder 9 8-6Ankle 8 7-6Calf 5 4-8Plantar arch 4 3-8Neck/cervical region 4 3-8Tibia 4 3-8Achilles tendon 3 2-9Thigh 2 19Heel 1 11st Metatarsophalangeal joint 1 ITotal 105 100

a Dichotomised groups based on average weeklytriathlon training load.Group 1: Swim .6 km, & bike .150 km, & .40 kmper week (n = 19).Group 2: All other subjects (n = 73).

136

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from

Low back pain in Japanese tniathletes

all Japanese. In comparison with other studiesinvolving triathlete subjects,3 5 6 the subjectswere generally younger (by approximately fiveto seven years), although their age was similarto that of the subjects in the study by Irelandand Micheli.2 Also, they were generally lighterand shorter. The majority of triathletes in thisstudy were distributed in the 26-30 and 31-35year age groups. The subjects trained some-what less than those in the previously men-tioned studies, but this was most probably dueto the inclusion of a wider range of triathletesin terms of skill, training, and competitionlevel. Almost all subjects followed a regularstretching programme, compared to the figureof 65% reported by Ireland and Micheli.Interestingly, very few subjects had occu-pations related to sport or triathlon, themajority being primarily involved in employ-ment of a non-physical nature.The methodological approach employed did

not allow generalisation about the entire popu-lation of Japanese triathletes. However, theresults were perhaps more descriptive of generaltriathlon populations than those in previousstudies, as the subjects were not selectedthrough participation in a particular competition(for example, full distance triathlon), but on thebasis of varying geographical location. Theincidence rates for all injuries may have beensomewhat exaggerated due to a possible non-participation by uninjured triathletes. However,we were aware of this problem from the out-set. We attempted to minimise this possibleselection bias by promoting the study as lookingat both triathlon training methods and triathlonoveruse injuries generally, rather thanemphasising the latter. Further, the question-naires were completed during the mid-seasontraining season, as opposed to after com-petitions. This could be expected to improve thereliability and generalisability of the results fortwo reasons. First, their training would havebeen of a more typical nature than that doneduring the competition season. Second, non-typical phenomena, such as normal post-racerecovery "pain" and fatigue, which could havebeen falsely interpreted as an injury, would havebeen excluded.

LOW BACK PAIN

The subjects in this study could be consideredto be much fitter and stronger than the generalpopulation, that is, all other persons notinvolved in regular and very physically de-manding athletic activities. The overwhelmingmajority did not smoke, were not overweight,and were most probably very determined andgoal oriented people, as reflected by theirinvolvement in the sport. Despite the relativelack of common low back pain risk factors, theincidence of low back pain of the subjects wasnot less than that reported for the generalpopulation, '0 as might be expected. While thepoint incidence (12%) was at the lower limit ofthe general population values, the lifetime incid-ence rate (59%) was in the middle range. Thisis interesting as the average age of the triathletesubjects was 31-3 years, with the age distributionof subjects in epidemiological studies involving

the general population ranging mainly from 30to 60 years.'0 In previous studies on the generalpopulation, either no gender differences hadbeen found, or the incidence had been greaterfor females, especially when work practices weretaken into account." In contrast, the males inthis study experienced a greater incidence oflowback pain than the females. The difference wasnot statistically significant, but was in fact veryclose (P = 0 052) to the set level of significance.This gender difference may be related todifferences in training methods or volumes. Forexample, differences for cycling and weighttraining between males and females were eitherstatistically significant or showed very strongtrends.Lumbar spine injuries account, on average,

for approximately 5-10% of all athleticinjuries, including both overuse and traumaticinjuries. 2 13 Sports such as gymnastics showthe highest low back injury rates.'4 1' Intriathlon, low back injuries account for 4l17%of all the overuse injuries sustainedlA 7 whichis comparable to the figures for other sports. Inthis study, low back complaints accounted foralmost 28% of all overuse injuries suffered inthe previous year. O'Toole et al5 reported thatan astonishing 72% of subjects had com-plained of low back pain or sciatica in theprevious year. There are very few data on theincidence rates of low back pain for cyclists,but Mellion'6 stated that up to 60% of cyclistshave suffered from it at some time. This closelymatches the life time low back pain incidencerate (59%) of the triathlete subjects in thisstudy. These figures show that low back painin triathletes is as prevalent as in other sports,and if traumatic low back injuries are excluded,then the prevalence would be even higher.According to Mooney's classification of low

back pain based on the potential for spon-taneous recovery,'7 the majority of low backpain in this group of triathletes could beattributed to soft tissue or myofascial structureinjuries. Involvement of the zygapophysealjoints was also likely, though a portion ofsuspected zygapophyseal joint problems couldalso represent a continuance of acute soft tissueinjuries due to secondary reflex activity, that is,the pain-spasm cycle. Of concern was thatalmost 20% of all the pain episodes had lastedover three months. According to the sameclassification, a more serious pathology orproblem may have existed, namely a possibleinjury of the intervertebral disc. While rela-tively frequent soft tissue injuries of shortduration would be expected for triathletesinvolved in regular, often very physically de-manding, training, the proportion of longduration pain episodes was quite high, and wasa clear danger sign of potential future andpossibly severe long term disability.

General risk factors for low back pain havebeen given for the general population.'o Itwould be expected that many of these riskfactors, especially the physical risk indicatorssuch as poor trunk muscle strength and endur-ance, cardiovascular fitness and flexibility, aswell as smoking, would be minimal in a groupof trained athletes. This study revealed that

137

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from

Manninen, Kallinen

increased trunk flexor muscle training fre-quency, that is, the number of times they weretrained weekly on average, was significantlyassociated with an increased incidence of lowback pain in the previous year. Possibly thetraining exercises for the trunk flexors wereperformed incorrectly. This may have resultedin inappropriate stresses on the lumbar spine.Also, it is possible that subjects began to traintheir trunk flexor muscles more frequently onlyafter they had experienced low back pain. Thismay have been done in an attempt to rehabili-tate themselves in the belief that greaterabdominal muscle training may decrease theirsymptoms. However, this may have actuallycontributed further to a deterioration or pro-longation of the acute phase symptoms.Another possibility is the presence of trunkmuscle imbalances. Low back pain subjectshave been seen to be flexor overpowered,"ratios of 1I0 to 0*8 (trunk extensors to flexors)having been reported.'9 20 Normal values are inthe order of 1 3.19 Perhaps the triathletesubjects with low back pain in this study wereflexor overpowered because they exercisedtheir abdominal muscles more than their backextensor muscles. However, it is not clearwhether an imbalance is a causal factor or aresult of low back pain. While it is very difficultto draw definite conclusions about the causes,clearly how the muscles are trained is just asimportant as how often and how much.

Other possible causes of low back pain inathletes have been suggested previously in theliterature.' 4 5 21 22 Excessive or repetitiveloading of the spine is a major cause of backinjury in endurance athletes.23 Repetitiveloading results in a weakening of the visco-elastic passive elements ofthe spinal structures.The loss of ability to protect these weakenedpassive elements makes the spine susceptible toinjury. 2 With an inability to dissipate forces atone joint, there is a resultant overload at thenext joint as forces ascend upwards towards theSkull.24 25 Should there be some kind of lowerlimb disorder, then abnormal forces could beincident on the lumbar spine.5 It has beensuggested that malalignment in the foot leadsto other foot problems, to knee problems, andsubsequently to lumbar spine problems.2' Thishas been seen in triathlete injury patterns, withback pain occurring frequently in combinationwith lower limb injuries.' 5While running may cause most overuse

injuries in triathletes,'13 this does not necess-arily mean it is the primary cause of low backpain. Cycling, may in fact be the keycomponent predisposing to low back pain,even though it is responsible for only 12-29%of overuse injuries in triathletes. 1 2 4 Migliorini4found that the change from the "off' load ofcycling, to the "on" load of running, as weightbearing begins, was the cause of most com-plaints. Muscle activity changes from beingmostly concentric during cycling to includealso eccentric muscle work during running. Ithas been postulated that a time lag exists aftercycling, before neuromuscular and elasticefficiency - which is indispensable for properrunning technique - can reach its optimal level.

The triathlete is perhaps more susceptible tolower back injury during the initial stages of therun, when this transition is occurring.Migliorini suggested that this transition periodlengthens with an increase in the length of thebiking phase. In our study, those experiencinglow back pain in the previous 12 monthsreported a greater average weekly cycling time,though this was not statistically significant(P = 0 065). This result provides some supportto Migliorini's suggestion. In order to maintainan aerodynamic cycling position, as well as anappropriate degree of hip flexion for optimalforce generation during the pedalling cycle, thelumbar spine is often held in a relatively flexedposition. Sitting, especially in a flexed positionand for long periods, has been shown toincrease intradisc pressure, thereby predispos-ing to low back pain.' 0 Thus it is very importantto examine cycling position and its associationwith the incidence of low back pain. Thepresent study did that, although very generally,with no significant associations being found. Itwould have been beyond the scope of this studyto examine the relation in greater detail. How-ever, future research should carried out to lookat how (inappropriate) cycling position maycause, or contribute to, low back pain. Further,the possible detrimental effects of vibration asa risk factor cannot be ruled out. So, whilerunning may have been the final factor produc-ing marked symptoms or injury, it is quitepossible that the initial cause of the low backpain originated from cycling. This is furthersupported by the majority of subjects them-selves implicating cycling as the main cause oftheir low back pain.There was some indication (P = 0 095) that

average weekly triathlon training load was asso-ciated with low back pain incidence. It isreasonable to assume that those training moreplace greater overall cumulative physical stresson their bodies than those who train less. Thiscould indicate that even among top leveltriathletes there is an upper "safe training limitor zone" beyond which an increase in trainingload would produce a non-proportional andgreater increase in the risk of injury. Of course,this theoretical upper limit would vary betweenindividuals, but perhaps certain generalisationscan be made.

OTHER INJURIESIn our study, the incidence of musculoskeletalinjuries related to triathlon was very similar tothose reported previously,' 2 4 with approxi-mately 60% of triathletes having suffered atleast one soft tissue overuse injury in theprevious year. Lower limb injuries accountedfor 61% of all injuries in this study, comparedto the figure of 85% reported by Ireland andMicheli.2 Also, the anatomical sites of injurywere closely matched. In all studies, the kneewas the single most common site of injury(30-50°/0). Injuries of the lower back (28%)were the second most common in this study,though in other studies they were the third mostcommon site (4-17%) after the shoulder,2 foot,or ankle.3 However, in the study by Collinset al,' the back was the most commonly injured

138

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from

Low back pain in Japanese triathletes 139

site. O'Toole et al5 revealed that low backinjuries occurred frequently in combinationwith lower limb injuries, although there were nostatistically significant associations in our study.The shoulder is the most common site ofupperlimb injuries, but the incidence has been quitevaried. In this study, injuries to the shoulderwere the third most common, and comprisedall upper limb injuries.

CONCLUSIONLow back pain is a common problem sufferedby triathletes. Despite the relative absence ofgeneral risk factors for low back pain,triathletes still suffer from it as much as thegeneral population or other athletes such ascyclists. While most pain episodes are of shortduration and involve the soft tissues, a rela-tively high proportion are of longer duration.These most probably implicate involvement orinjury of the intervertebral disc. This clearlyindicates that triathletes are at risk of sufferinglow back pain of a serious nature. Despite thefact that only one statistically significant differ-ence was found between low back pain andnon-low-back-pain subjects, there were severalother strong indications for possible asso-ciations of certain training related factors withthe incidence oflow back pain. The three mostcommon sites of injury in the present studywere of the knee, low back, and shoulder.While most knee injuries in triathlon arecaused by running, and the majority ofshoulder injuries by swimming, there is no oneclear cause of injury to the lower back, thoughcycling is suggested as a major contributingfactor. The sports related factors likely toincrease the risk oflow back pain must be moreclearly identified, and their effects minimised.The results of this study give direction for thedesign of controlled interventional studies, sothat these factors and their associations withlow back pain in triathletes can be investigatedmore accurately.

We would like to thank the staff at the Institute of SportsMedicine and Science in Japan, and the Daiko Foundation, forall their assistance, both organisational and financial, in relationto this study. Also, appreciation is extended to Tommi Salmelafor his assistance with statistical analysis, as well as to TiinaKuukkanen for her all her great organisational support.

1 Collins K, Wagner M, Peterson K, Storey M. Overuseinjuries in triathletes: a study of the 1986 SeafairTriathlon. Am_J Sports Med 1989;17:675-80.

2 Ireland M, Micheli L. Triathletes: biographic data, training,and injury patterns. Ann Sports Med 1987;3:117-120.

3 Massimino F, Armstrong M, O'Toole M, Hiller W, Laird R.Common triathlon injuries: special considerations formultisport training. Ann Sports Med 1988;4:82-6.

4 Migliorini S. An epidemiological study of overuse injuriesin Italian national triathletes in the period 1987-1990. JfSports TraumatolRelRes 1991;13:197-206.

5 O'Toole M, Hiller D, Smith R, Sisk T. Overuse injuries inultraendurance triathletes. Am Jf Sports Med 1989;17:514-8.

6 O'Toole M, Massimino M, Hiller D, Laird R. Medicalconsiderations in triathletes: the 1984 Hawaiian IronmanTriathlon. Ann Sports Med 1987;3:121-3.

7 Williams M, Hawley J, Black R, Freke M, Simms K. Injuriesamongst competitive triathletes. NZ J Sports Med1988;16:2-6.

8 Heliovaara M, Sievers K, Impivaara 0, Maatela J, Knekt P,MAkeld M, Aromaa A. Descriptive epidemiology andpublic health aspects of low back pain. Ann Med1989;21:327-33.

9 Kelsey J, White A. Epidemiology and the impact oflow backpain. Spine 1980;5:133-42.

10 Pope M. Risk indicators in low back pain. Ann Med1989;21:387-92.

11 Magora A. Investigation of the relation between low backpain and occupation. II. Work history. Ind Med Surg1970;39:504-10.

12 Harvey J, Tanner S. Low back pain in young athletes. Apractical approach. NZJ Sports Med 199 1;12:394-406.

13 Lachmann S. Soft tissue injuries in sport. London: BlackwellScientific Publications, 1988:3.

14 Goldstein J, Berger P, Windler G, Jackson D. Spine injuriesin gymnasts and swimmers: an epidemiological investi-gation. AmJ Sports Med 1991;19:463-8.

15 Meeusen R, Borms J. Gymnastic injuries. Sports Med1992;13:337-56.

16 Mellion M. Common cycling injuries. Sports Med 1991;11:52-70.

17 Mooney V. The classification of low back pain. Ann Med1989;21:321-5.

18 Pope M, Bevins T, Wilder D, Frymoyer J. The relationshipbetween anthropometric, postural, muscular, andmobility characteristics of males ages 18-55. Spine 1985;10:644-8.

19 Beenbon D, Morrisey M. A review of the literature relatedto trunk muscle performance. Spine 1988;13:655-60.

20 McNeill T, Warwick D, Anderson G, Schultz A. Trunkstrengths in attempted flexion, extension and lateralbending in healthy subjects and patients with low-backdisorders. Spine 1980;5:529-38.

21 Micheli L. Lower extremity overuse injuries. Acta MedScand (suppl) 1986;711:171-7.

22 Parniapour M, Nordin M, Kahanovitz N. Frankel V. 1988Volvo award in biomechanics. The triaxial coupling oftorque generation of trunk muscles during isometricexertions and the effect of fatiguing isoinertial movementson the motor output and movement patterns. Spine1988;13:982-92.

23 Mahlamaki S. Low back symptoms and signs in youngcross-country skiers: a cross-sectional and follow-up studyon the low back symptoms and signs, as well as relatedfactors, of 10-16 year skiers and controls living in theNorth-Savo area. Kuopio, Finland: Kuopio University,1987:23.

24 Voloshin A, Wosk J. An in vivo study of low back pain andshock absorption in the human locomotor system. JBiomech 1982;15:21-7.

25 Voloshin A, Wosk J, Brull M. Force transmission throughthe human locomotor system. J Biomech Eng 1981;103:48-50.

on January 28, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.30.2.134 on 1 June 1996. Dow

nloaded from