Stress Fractures of the Pelvis in Female Navy Recruits

Transcription

Stress Fractures of the Pelvis in Female Navy Recruits
MILITARY MEDICINE, 165,2:142,2000
Stress Fractures of the Pelvis in Female Navy Recruits:
Analysis of Possible Mechanisms of Injury
Guarantor: Edward W. Kelly, MD
Contributors: Edward W. Kelly, MD*; LCDR Scott R. Jonson, MSC USNt; Mark E. Cohen,
CDR Richard Shaffer, MSC USN§
The purpose of our study was to investigate possible risk
factors and mechanisms for the development of pelvic stress
fractures in female Navy recruits. We used a case-control retrospective study of female Navyrecruits undergoing basic military training. We compared anthropometric and activity data
between recruits with pelvic stress fractures (N = 25) and
female recruits who completed training without injury (N =
61). Recruits developing pelvic stress fractures were significantly (p < 0.05) shorter and lighter and were more frequently
Asian or Hispanic than recruits without stress fractures. In
addition, recruits with pelvic stress fractures reported marching in the back of their training division, were road guards, and
felt that their stride was too long during training activities
more often than recruits without injury. Self-reported fitness,
activities before recruit training, or a history of amenorrhea
was not found to be associated with the development of a
pelvic stress fracture in our population.
in similar activities.28-30 The proposed reasons for the
, , ,. ,. females' lower bone density. poor ---,--
"
with males.4.28.29.31
During our study,I
male recruits was diagnosedwith a pelvic stress
whereasthe femalerate was 1 in 367.
As ~ r~s~lt of thecprolongedhe~g time and ,-
population.Becausemilitary
-,-.-
compared
pelvic
stress
fractures
were shorterandc. had identifiable
r
r
~
c.
.
Introduction
S tress fractures have been a documented problem in the
Methods
military recruit for more than 140 years,I occurring in 1 to
30% of trainees.2-9Stress fractures have been estimated to cost Subjects
Volunteer
subjectsin this studywere
the military millions of dollars each year in lost training time,
,
.--'.
.
medical expenses, and trainee attrition. As a consequence of
such costs, a great deal of attention has been placed on identi- the Naval Recruit
through April I, 1996.
fying variables that may be associated with the developmentof the period March I,
tures
were
studied
in
females
exclusively because
stress fractures. Intrinsic variables may include fitness,IO,11'
incidence of pelvic stress fractures in females ,--bone density,12.l3bone width,14 bone morphology,II.13,15
and
-,.
-study period, 181 :
bone length.2,13Extrinsic variables may be related to the type,
.,--intensity, and duration of the activity performed.3,16-l8
Stress fractures most frequently affect the lower extremities,
(21%),pubic rami (14%),calcaneus
(7%),and
femur
with the tibia, metatarsals, calcaneus, and fibula being the most sal
,-,e
r
common bones involved.19,2O
Pelvic stress fractures, however,
pubic rami (N = 25).
are relatively rare, representing only 1 to 10% of all stress fractures reported.2l-25Stress fractures of the pelvis can be partic- completed recruit training free of any musculoskeletal injury
(N = 61). The study was approved by the Naval Hospital Comularly devastating because of the increased healing time remittee
for the Protection of Human Subjects, and informed conquired compared with stress fractures at other sites.26-28
The 6
sent
was
obtained from all participants. Participation was volto 12 months of convalescencerequired by many individuals
untary
and
without remuneration.
who suffer a pelvic stress fracture is much longer than the 4 to
The femalesin this study participated in 9 weeks of integrated
8 weeks of rehabilitation required for stress fractures at other
(males and females together) physical and military training.
sites.24
In the literature reporting pelvic stress fractures, females Each week, the recruits participated in formation marching,
have a much greater incidence compared with males engaging running, calisthenics, and military drill. To move around the
base, recruits marched in a formation of approximately 80 individuals, aligned from the tallest person in front to the shortest
.Resident In Orthopedic Surgery, MayoClinic, Rochester,MN 55905.
person in the back. Each division assigned 4 individuals to
tMuscuIoskeletalClinic, Naval Hospital, Great Lakes,IL 60088.
*Naval Dental ResearchInstitute, GreatLakes,IL 60088.
perform "road guard" duty that involved marching at the rear of
§NavalHealth ResearchCenter,San Diego,CA.
the training division. Road guards were subjected to greater
Presented,In part, at the Third Annual UnifornJedServicesRecruit Health Care
amounts of running because they were required to continually
Symposium,May 1997, Washington,DC.
race
ahead to halt traffic and then quickly rejoin the back of the
This manuscript was receivedfor reviewIn April 1999. The revisedmanuscript was
acceptedfor publication In August 1999.
training division before and after each intersection on the base.
Military Medicine,Vol. 165,February2000
142
143
of the Pelvis in Female Navy Recruits
At the time of diagnosis, subjects completed a 21-question
survey to record their age, height, weight, and known risk factors for low bone density. Questions assessed the subjects'
smoking history and history of prolonged amenorrhea (6 consecutive months or more without a period). Body mass index
was calculated as kg/m2. Each subject's self-perceived fitness
was assessedusing questions modified from the study of Washbum et al.32The subjects were asked, "Rate your physical fitness compared with others of your same age and sex," "How
many times a week did you run more than 1 mile during the 2
months before recruit training'?" and "In your leisure activities,
how often do you work up a good sweat?" For statistical purposes, the subjects were divided into two groups- "fit" or "unfit"-based on their answers to each of the three fitness questions. Subjects were considered unfit if they stated that they
werein "good,fair, or poor" condition, exercisedfewerthan three
times per week, and "rarely worked up a good sweat" while
exercising.
Additionally, subjects answered questions about activities
they participated in during recruit training, including "How frequently were you a road guard?" and "Where did you march in
the division?" Finally, subjects were asked about the length of
their stride while marching and running in formation: "Was your
stride too long, too short, or comfortable?"
Stress fracture diagnoseswere based on the recruit's subjective complaint of atraumatic, insidious onset of pain and tenderness consistent with positive radiographic findings (periosteal elevation, sclerosis, or fracture line) on plain film or bone
scintigraphy (focal area of increased uptake). All stress fracture
subjects were evaluated in the Naval Hospital's Musculoskeletal
Clinic. All radiographic findings were interpreted by the Naval
Hospital's staff radiologists. Control subjects had their medical
records reviewed to ensure that they had not been seen for a
musculoskeletal injury during recruit training.
DataAnalysis
The surveyresults weresubjectedto univariateanalysisusing the Tukeyrange,Kruskal-Wallis,or f testas appropriate.p
values < 0.05 were consideredsignificant.Analysiswas done
usingStatXact.
Results
Descriptive statistics for both groups are shown in Table I.
Table II summarizes the questionnaire results for the same
groups. The fitness questions are represented by the percentage
of subjects responding with answers that would indicate that
they were in the unfit category or "average or below average"
fitness as described by Washburn et al.32
In comparing the two groups, subjects developing pelvic
stress fractures were on averageshorter, lighter, and more often
Asian or Hispanic. but less often African American, than the
control group. In addition, subjects with pelvic stress fractures
marched in the rear of their training divisions. served as road
guards. and felt that their stride was too long while marching
and running more frequently than those subjects without injuries. Self-reported goodto poor fitness and a history of prolonged
amenorrhea were not found to be associated with the development of a pelvic stress fracture.
TABLE I
DESCRIIYrIVESTAllS'nCS BY GROUP
Group 1 (pelvic
Variable
Age(years)
Height(cm)
I Weight(kg)
Body~~s index
Ra~(o:
White
Afiican
American
I
Frac1:ures)
IN=, 25)
Group 2
(No
Injuries)
(N= 61)
Significance
I
21
159
55
22
36
4
70
20
32
2
7
24
Asian
~;-iSPaniC
4
1
Other
Dataare means(SO)or percentages.
A i statistic was computedfor
the race distribution,with Asianand Hispanicgroupedand compared
with all others(WhIte,Afiican American,Other).
.P< 0.01.
bp< 0.001.
Discussion
Pelvic stress fractures were fIrst described in military recruits
as early as 1937,23.33-35
each report citing only a few isolated
cases. More recently, Meunnan2o and Ozburn and Nichols3°
reported multiple cases of pelvic stress fractures in military
recruits. Pelvic stress fractures have also been reported in civilian athletes, most often afflicting long-distance runners26.28.36-36
and fencers.39
The cause of a pelvic stress fracture has beenattributed to the
pull of the adductor muscles on the narrow pubic rami.23.28.36
The muscular pull theory is supported by the fact that stress
fractures of the pelvis most frequently occur at the isthmus (Fig.
I), which is the narrowest area of the pubic bone (Fig. 2) as well
as the site for the origin of the adductor magnus.40During the
nonnal gait cycle, the medial portion of the adductor magnus
acts with the other adductor muscles to flex the hip, whereas its
lateral fibers act with the hamstrings to extend the hip.23.40
Selakovichand Love proposed that it was the interplay between
the adductor muscles, which originate medial on the pubic
rami, and the hamstrings, which originate laterally, acting as
antagonists during gait, that create a shear stress at the narrow
isthmus and produce a fracture.34
The results from our study clearly support this hypothesized
mechanism. Those females who developed stress fractures of
the pelvis tended to be shorter than average,marched at the rear
of the training division, were frequently road guards, and complained the most of having to overstride while marching. These
findings fit into a consistent explanatory framework for pelvic
stress fractures driven by height and the assignment of individuals, by height, into marching fonnations. Short recruits are
assigned to the rear of marching fonnations. Here, these short
individuals are required to overstride to keeppace with the taller
fonnation leaders. When the short individuals are forced to take
"giant steps. and overstride, the accumulation of the large shear
and/or tensile stresses placed on the pubic rami by the adductors and hamstrings is amplified.
Military Medicine.Vol. 165. February 2000
(",'.
~
144
Stress Fractures of the Pelvis in Female Navy
TABLE UQUES110NNAIRE
RESULTS BY GROUP
Data are percentages or means (SD).
ap< 0.001.
bp< 0.01.
~
(y
1..""
\
'~r'
II
/
y
Fig. 1. Radiographdemonstrating the typical location and appearanceof a pelvic
stress fracture.
Ozburn and Nichols noted that many of the female Amy
recruits suffering pelvic stress fractures were short and also
marched at the back of the company. 30The recruits commented
on having to "run or take giant steps to keep up" with the taller
males at the front of their training division. Hill and colleagues
were able to eliminate the occurrence of pelvic stress fractures
in female British Amy recruits by decreasingthe stride length of
a marching training regiment to 27 inches from the standard
30-inch pace.41In another study. Reinker and Ozbume were
able to decreasethe incidence of all stress fractures in a female
Amy trainee population from 12% to 1.5% by switching the
marching order from progressively shorter to progressively taller.31Our study is the first we have identified that demonstrates
statistically that the combination of being short and marching at
Military Medicine,Vol. 165, February 2000
Fracture
site
Adductor
magnus
Hamstrings
Fig. 2. The origin of the adductor magnus and hamstring muscles. shown with
the most common site of pelvic stress fracture.
the rear of a tall-to-short aligned division significantly contributes to the development of pelvic stress fractures in female
recruits.
Other authors have noted a similar relationship between
shorter female height and injury rates among military recruits.4,42Giladi et al.43and Finestone et al.2suggestedthat it is
the short, narrow bones that are at risk for stress fracture.
Barrow and Saba. however, found no relationship between
height and stress fracture risk among female athletes.12In an
Stress Fractures of the Pelvis in Female Navy Recruits
145
anthropometric study of 626 male U.S. Marine Corps recruits, factors for subjects with symptomatic stress fractures. In addiBeck and colleagues concluded that recruits who developed tion, we did not perform specific quantitative measures of bone
density, bone morphology, or skeletal alignments, which would
stress fractures were shorter, lighter, and had smaller anthropometric dimensions than recruits who were injury free.13They give more specific information. Nonetheless. we feel that our
did not. however, differentiate by site of fracture. nor did they findings shed light on the mechanism of injury for pelvic stress
account for possible differences in the activities of recruits who fractures and open the door to further research in this area.
In June 1996, several changes were made in an attempt to
did or did not developstress fractures.
decrease
these injuries in our training population. By having the
The contradictions about short stature as an intrinsic risk
factor for stress fracture in general may be explained by our shortest individuals in a division march in front, prohibiting the
findings. We believe that short stature is not an independent road guards from running. and rotating the road guard position
on a daily basis, the overall stress fracture rate in our female
risk factor for pelvic stress fractures when individuals are alrecruits decreasedfrom 3.5% to 1.3%. During this same period.
lowed to walk, march, or run at their natural pace and stride.
the pelvic stress fracture rate decreased by more than 80%.
This is the case. for example, with athletes who run at a natural
These results support our conclusions that overstriding is the
stride length. Short stature doesbecome a risk for pelvic stress
major source of stress for the development of pelvic stress fracfractures, however. when an individual is forced to .overstride,. tures in female Navy recruits.
such as when marching in a formation of recruits aligned from
tall to short.
Conclusions
In our study population. 20% of the control subjects in group
2 were African American, whereas less than 4% of the subjects
Our study supports the theory that stress fractures at differwho developedpelvic stress fractures were African American. A ent anatomic sites may result from different mechanisms. The
low rate of stress fractures in African-American subjects has most significant risk factor in our population for the developbeen reported by several other investigators in both military
ment of a pelvic stress fracture is the combination of being short
and athletic populations.s.1o.12
The apparent protective factor of and marching at the rear of an integrated training division. The
African-American race against stress fractures may be attriboverstriding was easily altered, decreasing the incidence of all
uted to the demonstrably higher bone density of African-Amer- stress fractures in our female recruits. The association between
ican individuals compared with whites. Asians. and Hispan- Asian or Hispanic race and the development of pelvic stress
ics.44.45
fractures is likely related to the shorter average height in these
The majority of the women developing pelvic stress fractures individuals, but this may require further review.
in the present study were Asian (24%), Hispanic (32%), or white
(36%).The increased proportion of Hispanic and Asian females
References
in the pelvic stress fracture group may be attributable to the
shorter-than-average height in these ethnic populations or to
1. Brlethaupt MD: Zurpathologie des menschliched fusses. Med Zeltung1855; 24;
169.171. 175-7.
differences in anthropometric factors. such as pelvic bone ge2. FlnestoneA. Shlamkovitch N. Eldad A: RIsk factors for stress fractures among
ometry or body mass index. To the best of our knowledge, no
Israeli tnfanny recruits. MllIt Med 1991; 156: 528-30.
other published reports demonstrate a similar predilection of 3. Jones BH. CowneOM. Knapik JJ: Exercise. training and injuries. Sports Med
pelvic stress fractures in Asian or Hispanic individuals. Further
1994; 18: 202-14.
4. KowalNJ: Nature and causesof injuries In women resulting from an endurance
research is needed in this area.
training program. Am J Sports Med 1980; 8: 265-9.
Increasing age,S.IO.46
cigarette smoking,IO and extremes of
5. MilgramC. Giladi M. Stelrn M: Stressfractures In mIlItmy recruits: a prospective
body mass4.42
have all been demonstrated in other studies to be
study showing an unusually high incidence. J Bone Joint Surg Br 1985; 67:
predictive for the development of overuse injuries, stress frac732-5.
tures in particular. These variables, as well as other predictors
6. Morris JM. Blickenslaff JD; FatigueFractures: A Clinical Study. Springfield. IL.
Charles CThomas. 1967.
for low bone density, such as prolonged exercise-induced am7. PesterS. Smith PC: Stress fractures In the lower extremities of soldiers In basic
enorrhea12.IS.47
and poor fitness,46did not appear to increase the
risk for stress fractures in our population. The lack of a correlation between factors predictive of low bone density and the
development of pelvic stress fractures indicates that the main
risk factor for such injuries in our population is activity related
rather than an intrinsic lack of fitness.
The relatively small number of subjects in each group limited
our study. Our study was also limited by data collected retrospectively through a survey. We relied on self-reported heights
and weight for subjects in both groups, which may have been
inaccurate. Becausethe survey was conducted at the time of the
injury and the activity a recruit is exposedto during training is
fairly uniform, we are confident of the reliability of the participants' answers. We did not make radiographs or bone scans of
individuals in the control group that might have identified
asymptomatic stress fractures. We do not feel that this is relevant, however,because the goal of our study was to identify risk
training. Orthop Rev 1992; 21: 297-303.
8. Schmidt BrudvigTJ. GudgerTD. ObermeyerL: Stress fractures In 295 trainees:
a oneyear study of incidenceas related toage. sexand race. MIlIt Med 1983; 148:
666-7.
9. ScullyTJ. BestermanG: Stressfracture: a preventabletraining Injury. MIlIt Med
1982; 147: 285-7.
10. FrIedt KE. NuovoJA. PatienceTH: Factors assocIatedwith stress fracture In
young Army women: Indications for further research. MIlIt Med 1992; 157:
334-8.
11. Giladl M. Milgram C. Sin1kInA: Stressfractures: Identifiable risk factors. Am J
Sports Med 1991; 19: 647-52.
12. Barrow GW. Saba S: Menstrual irregularIty and stress fractures In collegiate
femaledislance runners. Am J Sports Med 1988; 16: 209-16.
13. BeckTJ. RuffCB. Mourtada FA: Dual-energyX-ray absorptiomenyderivedstructural geomenyfor stress fracture prediction In maleU.S. Marine Corpsrecruits.
J Bone MIner Res 1996; 11: 645-53.
14. Giladl M. Milgrom C. Sin1kInA: Stress fractures and tibial bonewidth. J Bone
Joint Surg Br 1986; 69: 326-9.
15. Giladi M. Milgrom C. Stein M: External rotation of the hip: a predictor of risk for
stressfractures. Clin Orthop Relat Res 1987;216: 131-4.
Military Medicine.Vol. 165. February2000
16. BennellKL. ThomasSA,.,
track and field athletes. Am J Sports Med 1996; 24; 211-7.
17.
track-and-field athletes: a retrospectiveanalysis. Clin J Sport Med
229-35.
18. KadelNJ. Teltz CC. Kronmal RA: Stressfractures In ballet dancers.!
Med 1992; 20: 445-9.
19. Wilson ES. Katz FN: Stress fractures: an analysis of 250 consecutive
Radiology1969;92: 480-6.
20. Meunnan KO: Stress fracture of !he pubic arch In militmy recruits. ]
1980; 53: 521-4.
21. DevasMB: StressFractures. EdInburgh. Churchill LIvingstone. 1975.
22. ,
---,
, -,'
"-- -,
~ .Radiology1943;41; 586-8.
23. LevetonAL: March fractures of the long bones of the lower extremity and pelvis.
AmJSurg 1946;71: 222-32.
24. Matheson GO, Clement DB, McKenzieJE, et al: Stress fractures In athletes: a
study of 320 cases.Am J Sports Med 1987; 15: 46-58.
25. McBrydeAM Jr: Stress fractures In runners. Clin Sports Med 1985;4: 737-52.
26. Noakesm, Smith JA, Undenberg G: Pelvic stress fractures In long distance
runners. Am J Sports Med 1985; 13: 120-3.
27. O'Brten T, Wilcox N, Kersch T: Refractory pelvic stress fracture In a female
long-distancerunner. Am J Orthop 1995; 24: 710-3.
28. PavlovH, Nelson'fL. WarrenRF: Stressfractures of the pubtc ramus: a report of
twelvecases.J Bone Joint Surg Am 1982; 64: 1020-5.
29. GoldbergB, PecoraC: Stressfractures: a rtsk of Increasedtraining In freshman.
PhystclanSportsmed 1994;22: 68-78.
30. OzburnMS. NicholsJW: Pubic ramus and adductor Insertion stress fractures In
femalebasic trainees. Mllit Med 1981; 146: 332--4.
31. RelnkerKA. OzburneS: A compartsonof male and femaleorthopedicpathologyIn
basic training. Mllit Med 1979: 144: 532-6.
32. WashburnRA. AdamsLL, HalleGT: PhysicalactMtyassessmentfor epidemiologic
research:the utility of two slmp]jftedapproaches.PrevMed 1987; 16: 636-46.
33. Nickerson H: March fracture or insufficiency fracture. Am J Surg 1943; 62:
154-64.
34. SelakovlchW, LoveL: Stressfractures of the pubic ramus. J Bone Joint SurgAm
1954; 36: 573-6.
35. WachsmuthW: Zur atlologic der schlelchendenfrakturen. Der Chlrung 1937;9;
16-24.
36. LatshawRF, Kantner TR. KalenakA; A pelvic stress fracture In a femalejogger: a
casereport. Am J Sports Med 1981; 9: 54-6.
37. OravaS. Puranen J, Ala-KetolaL: Stress fractures causedby physical exercise.
Acta Orthop Scand 1978;49: 19-27.
38. Thorne DA. FredertckDL: Pelvicstress fracture In femalerunners. Clin Nucl Med
1986; II: 828-9.
39. Forcher-Mayr0: Beitragzu den traintngveranderungenbelm sportfechten. Wlen
KIln Wochenschr1954; 63: 331-3.
40. Moore KL: Clinically ortented anatomy [3rd Ed.). Baltimore, MD, Williams and
Wilkins, 1992.
41. Hill PF, Chatlerjl S, ChambersD; Stress fracture of the pubic ramus In female
recruits. J Bone Joint Surg Br 1996; 78: 383-6.
42. Jones BH, BoveeMW, HarrtsJ: Intrinsic rtsk factors for exercise-relatedInjurtes
amongmale and femaleArmy trainees. Am J Sports Med 1993;21: 705-10.
43. Gliadl M. Mligrom C, Simkin A; An analysis of the blomechanlcalmechanism of
tibial stress fractures among Israeli Infantry recruits. Clin Orthop Relat Res
1991; 231: 216-22.
44. Trotter M. Broman GE. PetersonRR; Densities of bones of whIte and Negro
skeletons.J Bone Joint Surg Am 1960;42: 50-8.
45. Cohn SH, AbesemisC, Yosumura S: Comparativeskeletal mass and radial bone
mtneral content In black and white women. Metabolism 1977;26: 171-8.
46. Macera CA, KIrby JL, HagenmaierGW: Age. physical activity. physical fitness,
body composition. and incidence of orthopedicproblems. Res Q ExerciseSport
1989; 60: 225-33.
47. Drinkwater BL, Nilson K, ChestnetCH: Bone mtneral content and amenorrheic
and eumenorrhelcathletes. N Engi J Med 1984:311: 277-81.
48. Block JE, FrtedlanderAL. Brooks GA; Determtnants of bone density among
athletes engaged In weight-bearing and non-weight-bearing activity. J Appl
Physlo11996;24: 211-7.
Military Medicine.Vol. 165. February 2000