Fatigue Bone Stress Injuries of the Lower Extremities in Finnish Conscripts

Home , Talus bone

From the Department of Radiology, University of Helsinki; and the Research Institute of Military Medicine, Central Military Hospital, Helsinki

Fatigue bone stress injuries of the lower extremities in Finnish conscripts

Maria Niva

Academic dissertation

To be presented, with the assent of the Faculty of Medicine of the University of Helsinki, for public discussion in Small hall of the University main building on February 10th, 2006 at 12 noon.

Helsinki 2006 Supervised by

Docent Harri Pihlajamäki, M.D., Ph.D. Department of Surgery and Research Institute of Military Medicine Central Military Hospital Helsinki, Finland and

Docent Martti Kiuru, M.D., Ph.D, MSc. Department of Radiology Helsinki University Helsinki, Finland

Reviewed by

Docent Jari Parkkari, M.D., Ph.D. Tampere University Tampere, Finland and

Docent Riitta Parkkola, M.D., Ph.D. Department of Radiology Turku University Turku, Finland

To be discussed with

Docent Osmo Tervonen, M.D., Ph.D. Department of Radiology Oulu University Oulu, Finland

ISBN 952-91-9857-4 (nid.) ISBN 952-10-2923-4 (PDF) Press: Yliopistopaino, Helsinki 2006 To all who are interested in bone stress injuries Contents

ABSTRACT ...... 6

LIST OF ORIGINAL PUBLICATIONS ...... 8

ABBREVIATIONS ...... 9

1. INTRODUCTION ...... 10

2. REVIEW OF THE LITERATURE ...... 12 2.1. Historical aspects of bone stress injuries ...... 12 2.2. Terminology of bone stress injuries ...... 13 2.3. Bone marrow, epiphyses, metaphyses ...... 16 2.4. Bone reaction to stress ...... 18 2.5. Incidence of bone stress injuries ...... 20 2.6. Etiology of bone stress injuries ...... 20 2.7. Clinical diagnosis of bone stress injuries ...... 22 2.8. Diagnostic imaging of bone stress injuries ...... 23 2.8.1. Radiography ...... 23 2.8.2. Scintigraphy ...... 23 2.8.3. Magnetic Resonance Imaging ...... 24 2.9. Differential diagnosis of bone stress injuries ...... 25 2.10. Asymptomatic bone stress injuries ...... 28 2.11. Symptomatic bone stress injuries ...... 28 2.12. Bone stress injuries in the lower extremity ...... 29 2.12.1. Bone stress injuries of the hip ...... 30 2.12.2. Bone stress injuries of the femoral shaft ...... 30 2.12.3. Bone stress injuries of the knee ...... 31 2.12.4. Bone stress injuries of the ankle and foot ...... 32

3. AIMS OF THE PRESENT STUDY ...... 35

4 4. MATERIALS AND METHODS ...... 36 4.1. Patients ...... 36 4.2. Clinical diagnosis ...... 39 4.3. Imaging methods ...... 39 4.3.1. Pelvis and hip ...... 40 4.3.2. Knee ...... 40 4.3.3. Ankle and foot ...... 40 4.4. Image interpretation ...... 41 4.5. Statistical methods ...... 41

5. RESULTS ...... 44 5.1. The occurrence and clinical signiﬁ cance of asymptomatic bone stress injuries in elite recruits (I) ...... 44 5.2. The incidence, location, and nature of bone stress injuries in the knee (II) ...... 44 5.3. The incidence, location, and nature of bone stress injuries in the femoral bone (III) ...... 48 5.4. The incidence, location, and nature of bone stress injuries in the ankle and foot (IV) ...... 54

6. DISCUSSION ...... 63 6.1. Patients and study design ...... 63 6.2. The occurrence and clinical signiﬁ cance of asymptomatic bone stress injuries ...... 63 6.3. Incidence of bone stress injuries ...... 64 6.4. Location and nature of bone stress injuries ...... 65 6.5. Imaging of bone stress injuries ...... 67

7. CONCLUSIONS ...... 68

8. SUMMARY ...... 70

9. ACKNOWLEDGEMENTS ...... 73

10. REFERENCES ...... 75

ORIGINAL PUBLICATIONS ...... 84

5 Abstract

Bone stress injuries resulting from overuse are frequent not only in athletes and in military trainees but also among healthy people who have recently started a new or intensive physical activity. Diagnosis of bone stress injury is based on the patient’s history of increased physical activity and on imaging fi ndings. In the prospective study I, the occurrence and the clinical signifi cance of asymptomatic bone stress injuries in elite-unit military recruits undergoing intensive physical training in the Parachute School of the Utti Jaeger Regiment was examined by clinical and MRI follow-up before commencement of the intensive training period, at 6 weeks from commencement, and on completion of the 5-month training program. In the retrospective studies II-IV, the incidence, location, nature, and patterns of bone stress injuries in the knee, the femoral bone, and the ankle and foot were examined based on MRI. In all studies I-IV, the MR images with bone stress injury fi ndings were evaluated with respect to bone stress injury location and bone stress injury type, and bone stress injuries were graded. The results of the study I indicate that 16 of a total of 21 recruits displayed altogether 75 bone stress injuries on three MRI scans. Only 40% of the bone stress injuries were symptomatic. Symptoms depended on the location and MRI grade of injury, with higher grades usually more symptomatic. Repeated clinical and MRI assessment indicated that 84% of asymptomatic grade I bone stress injuries healed or remained as grade I and asymptomatic. In the study II, the incidence of bone stress injuries in the knee was 103 per 100,000 person-years (88 patients with a positive MR fi nding out of 85,318 conscript-years, 141 bone stress injuries, a total of 1330 patients MR imaged). Bilateral bone stress injuries of the knee were found in 25%, and two solitary bone stress injuries in the same knee simultaneously in 28% of the patients. The most common anatomic location was the medial tibial plateau, which was also the most typical location for a more advanced injury. After the commencement of the military service,

6 a bone stress injury in the medial tibial plateau caused knee pain earlier than a bone stress injury elsewhere in the knee (p=0.014). In the study III, the incidence of femoral bone stress injuries was 199 per 100,000 person-years (170 patients with a positive MR fi nding out of 85,318 conscript-years, 185 bone stress injuries, a total of 1857 patients MR imaged). The three most common sites were the neck (50%), the condylar area (24%), and the proximal shaft (18%). A stress reaction occurred in 57%, and a fracture line in 22% of the cases. There was a statistical correlation between the bone stress injury grade and the duration of pain (p=0.001). Bone stress injuries in the upper femur showed a shorter duration of pain than bone stress injuries in the condylar area of the knee. In the study IV, the incidence of bone stress injuries in the ankle and foot was 126 per 100,000 person-years (131 patients with a positive MR fi nding out of 104,340 conscript-years, 378 bone stress injuries, a total of 267 patients MR imaged). 58% of the injuries occurred in the tarsal bones and 36% in the metatarsal bones. Multiple bone stress injuries in one foot occurred in 63% of the cases. In tarsal bones except the calcaneus and in metatarsal bones except the fi fth metatarsal bone, a bone stress injury was signifi cantly more often seen in more than one bone. In conclusion, asymptomatic grade I bone stress injuries are common in subjects undergoing intensive physical training and they heal or remain asymptomatic grade I bone stress injuries even when the intensive physical activity continues. They are, therefore, of no clinical signifi cance. The incidence of bone stress injuries in the knee underlying exercise- induced knee pain seems to be relatively high in conscripts. In the knee, the most common anatomic location and the most typical location for a more advanced injury was the medial tibial plateau. Femoral bone stress injuries seem to be relatively common in physically active conscripts. In the femur, the most common anatomic site was the neck followed by the condylar area and the proximal shaft. The location of pain was usually situated in the site of the femoral bone stress injury. Multiple various- stage bone stress injuries of the ankle and foot were found in physically active young adults experiencing exercise-induced ankle and/or foot pain during military service. The most commonly affected anatomic area were the tarsal bones. As the treatment and treatment period of bone stress injuries vary depending on the bone involved and the injury grade, importance of an early diagnosis by MRI should be emphasized for physically active young adults suffering from stress related lower extremity pain.

7 List of original publications

This thesis is based on the following articles, which are referred to in the text by their Roman numeral:

I Kiuru MJ, Niva M, Reponen A, Pihlajamäki HK: Bone stress injuries in asymptomatic elite recruits: a clinical and magnetic resonance imaging study. Am J Sports Med 2005; 33(2):272-276.

II Niva MH, Kiuru MJ, Haataja R, Pihlajamäki HK: Bone stress injuries causing exercise-induced knee pain. Am J Sports Med 2005; Sep 16.

III Niva MH, Kiuru MJ, Haataja R, Pihlajamäki HK: Fatigue injuries of the femur. J Bone Joint Surg Br 2005; 87(10):1385-1390.

IV Niva MH, Sormaala MJ, Kiuru MJ, Haataja R, Ahovuo JA, Pihlajamäki HK. Bone stress injuries of the ankle and foot, an MRI study, submit- ted.

The publishers have kindly granted permission to reprint the original articles.

8 Abbreviations

99mTC technetium-99m

CT computerized tomography

F female

FOV ﬁ eld of view

FSE fast spin echo

GE gradient echo

L left

M male

MR magnetic resonance

MRI magnetic resonance imaging

N number

P probability

R right

SE spin echo

STIR short tau inversion recovery

T tesla

T1 longitudinal relaxation time

T2 transverse relaxation time

TE echo time

TI inversion time

TR repetition time

9 1. Introduction

The fatigue bone stress injuries involving the lower extremities play an essential role not only on the individual level among military conscripts and athletes, but also as a public health concern. In the Finnish Defence Forces, the incidence of bone stress injuries has varied from 1.0% to 24.5% depending on the military branch and training program (Sahi et al. 1996). During the years 1999-2004, musculoskeletal conditions and musculoskeletal trauma accounted for mean 20% of all service interruptions during the military service. Diagnosis of a bone stress injury is based on a patient´s history of increased physical activity, on physical examination, and on imaging fi ndings (Anderson and Greenspan 1996). In MRI, the nearly 100% sensitivity is the same than that of bone scintigraphy (Stafford et al. 1986, Matheson et al. 1987), but its specifi city is superior to that of scintigraphy (Daffner 1978, Stafford et al. 1986, Lee and Yao 1988, Meyers and Wiener 1991, Shin 1996, Kiuru et al. 2002). Over recent years, even though bone scintigraphy is used as the golden standard in the diagnosis of bone stress injuries (Daffner 1978, Anderson and Greenspan 1996), MR imaging has become one of the fi rst-choice diagnostic modalities for stress-related bone changes (Stafford et al. 1986, Lee and Yao 1988, Fredericson et al. 1995, Anderson and Greenspan 1996, Arendt and Griffi ths 1997, Boden et al. 2001, Kiuru et al. 2002). MRI is the most specifi c technique for the detection of bone marrow edema, periosteal edema, surrounding soft tissue edema, fracture line, and callus (Bergman and Fredericson 1999, Kiuru et al. 2001). In the Central Military Hospital, Helsinki, Finland, MRI is widely used for diagnosis of bone stress injuries and musculoskeletal conditions in patients with symptoms but eliciting negative radiographic images. Successful management of bone stress injuries among military conscripts and athletes requires an understanding of the type of the physical activity, precise knowledge of the anatomy of the bones, and aware- ness of bone stress injuries. Most fatigue bone stress injuries respond well to a combination of conservative therapies. Typically, bone stress

10 injuries tend to appear during the two- to three-month long basic training period demanding an abrupt increase in the intensity or duration of physical training of the conscripts, to which they are often not accustomed in the civilian life (Morris and Blickenstaff 1967, Provost and Morris 1969, Geslien et al. 1976, Engber 1977, Milgrom et al. 1985, Mil- grom et al. 1986, Muralikuttan and Sankarart-Kutty 1999, Välimäki et al. 2005). Bone stress injuries are specifi c both in location and in regard to the activity that produced them (Daffner 1978). The risk for development of bone stress injuries is exceedingly common in special elite groups (Sahi 1984). Even though asymptomatic bone stress injuries have been described to occur after physical exercise, their clinical meaning is unclear (Kiuru et al. 2003, Hohmann et al. 2004). The purpose of the fi rst prospective cohort study (I) was to evaluate by clinical and MRI follow-up the occurrence of asymptomatic bone stress injuries, their clinical signifi cance, and whether they progress to stress fractures in elite-unit military recruits undergoing intensive physical training. Mainly case reports about stress fractures of the knee area have been published so far (Geslien et al. 1976, Hensal et al. 1983, Milgrom et al. 1986, Weber 1988, Orava et al. 1996, Boden and Speer 1997, Yasuda et al. 1992, Glorioso et al. 2002). The previous studies on femoral bone stress injuries mostly deal with the femoral neck and the proximal femoral shaft (Egol et al. 1998, Kiuru et al. 2003), whereas bone stress injuries of the femoral head, the middle-distal shaft, and the condylar area are reportedly uncommon (Visuri 1997, Salminen et al. 2003). Multiple bone stress injuries are common in the foot area (Meurman and Elfving 1980a). The purpose of the retrospective studies (II-IV) was to assess the incidence, location, nature and patterns of bone stress injuries based on MRI in the knee area, the femur, and the ankle and foot in physically active young adults who reported of exercise-induced lower-extremity pain disturbing their military service.

11 2. Review of the literature

2.1. Historical aspects of bone stress injuries

In 1855, stress fractures were fi rst mentioned by J. Breithaupt (Breith- aupt 1855), a German military surgeon who described painful swollen feet in young military recruits. Since there were no radiographs avail- able, he failed to recognize the spontaneous metatarsal fracture, but distinguished this phenomenon from strain of the metatarsals and from Morton´s neuroma, a condition which he himself had. He named the condition “Fussgeschwulst” (foot swelling or foot tumor), because he believed that the swelling was caused by a traumatic infl ammatory reaction in tendon sheats (Morris and Blickenstaff 1967). After Wilhelm Röntgen had discovered X-rays in 1895, the fi rst ra- diograph of a stress fracture, called a “march fracture”, of a metatarsal shaft was published by Stechow in 1897 (Stechow 1897). As radiographs became more widely used, a considerable experience with fatigue fractures occurring also in bones other than the metatarsals was acquired in military basic training centers. In 1905, Blecher wrote the fi rst report of a fatigue fracture of the femoral neck (Blecher 1905). In 1936, Asal described a displaced femoral neck fracture in a German army recruit, as well as bilateral undisplaced femoral neck fractures occurring after long marches with heavy loads (Asal 1936). In 1929, Ollonqvist published the fi rst radiographs of periosteal fatigue callus in the tibia of Finnish conscripts (Ollonqvist 1929). He suggested the term “osteopathia itineraria” for this disorder, deriving the name from the Latin iter, itineris “a march.” The name referred to the etiology of the disease. Also in 1929, Aleman called attention to fatigue fractures of the tibia and reported that approximately 100 cases of tibial fatigue fractures occurred each year in the Swedish army (Aleman 1929). Genz (1935) and Asal (1936, 1937) in Germany and Ollonqvist (1937) in

12 Finland described that fatigue fractures of the tibia occurred during the early phases of the military training (Genz 1935, Asal 1936, Asal 1937, Ollonqvist 1937). A fatigue fracture of the tibial and femoral shaft has been regularly misdiagnosed as a malignant bone tumor leading either to bone biopsy or, at worst, “trial osteotomy” (Asal 1936, Morris and Blickenstaff 1967). In imaging studies, healing-phase bone stress injury may show characteristics typical of malignant tumors. In the histologic studies about healing-phase bone stress injury, differentiation of benign and malignant changes is challenging, even impossible. In the setting of a potential fatigue bone stress injury or a malignant tumor, a biopsy should be avoided since immature bone cells and osteoid of a bone stress injury representing the healing fracture process can be misinterpreted for a malignancy (Devas 1975, Daffner and Pavlov 1992). After development of scintigraphy and MRI (Damadian 1971, Geslien et al. 1976, Damadian et al. 1977), it has been possible to describe bone stress injuries in varying anatomic locations. Further development has introduced a number of MRI grading scales for the assessment of different MRI signs (Lee and Yao 1988, Fredericson et al. 1995, Kiuru et al. 2001). Terms osteopathia itineraria, soldier´s fracture, insidious fracture, crack fracture, pseudofracture, spontaneous fracture, exhaustion fracture, overload fracture, fatigue fracture, and stress fracture have been used to describe this bone injury (Jones et al. 1989). Nowadays, with the occurrence of fatigue bone stress injuries in athletes and children being well-known (Orava et al. 1978, Peltokallio 2003), terminology varies according to the degree of mineral or elastic resistance of the bone (fatigue-insufﬁ ciency), the type of stress exerted on the bone (compression-tension), and the different risks for complications (low risk – high risk).

2.2. Terminology of bone stress injuries

Consistent and adequate deﬁ nitions of terminology are necessary to permit comparisons between the existing studies. Stress reaction starts when an abnormal amount of stress is exerted on a bone without an adequate rest period to allow for adaptation. Bone stress reaction is visualized as an area of prefracture bone remodeling in which the bone is weakened but not physically disrupted (Boden

13 and Osbahr 2000). No evidence of a fracture line is indicated (Jones et al. 1989). Depending on the phase of the physiological spectrum in MRI, marrow edema is seen as an increased signal on STIR and fat-suppressed T2-weighted fast spin echo sequences, and as a decreased signal intensity on T1-weighted spin echo sequences (Stafford et al. 1986, Vogler and Murphy 1988, Lee and Yao 1988). A marrow edema seen on MR images may represent interstitial fl uid (increased accumulation of extracellular fl uid), trabecular microfractures and/or blood at histologic examination (Vogler and Murphy 1988, Fazzalari 1993, Li et al. 1998). Scintigraphy shows foci of increased activity, and radiographs are negative. Fatigue stress fracture occurs when abnormal stress, usually in the form of multiple and frequent repetition of otherwise normal stress, is exerted on a normal bone with normal elastic resistance but unac- customed to that exertion (Daffner 1978). MRI shows a cortical or an intramedullary low-signal band or zone on T1- and T2-weighted images (Lee and Yao 1988). Callus becomes apparent as an intermediate cortical signal intensity (Stafford et al. 1986, Lee and Yao 1988, Vogler and Murphy 1988, Meyers and Wiener 1991, Fredericson et al. 1995, Ander- son and Greenspan 1996). Immature bone cells may be present in case of a healing stress fracture (Devas 1975). Scintigraphy shows a “fracture line” with focal, cortical, or medullary zone of increased radionuclide uptake. Insuffi ciency stress fracture results when normal or physiologic stress such as stretching, compression, bending, or torsion is exerted on a bone with a low mineral content (Pentecost et al. 1964, Daffner 1978). Insuf- fi ciency fracture has been reported to be associated with osteoporosis, rheumatoid arthritis, psoriatic arthritis, juvenile chronic arthritis, anky- losing spondylitis, diabetes mellitus, hyperparathyroidism, pseudohy- perparathyroidism, Paget disease, chronic liver disease, chronic renal disease, anemia, amenorrhea, neoplasms, decreased testosterone level, glycogen storage diseases, cerebral palsy, corticosteroid therapy, osteo- genesis imperfecta, eating disorders, rickets, and other nutritional defi - ciencies (Pentecost et al. 1964, Mäenpää et al. 2002). Pathological fracture occurs through a region of bone that has been weakened by a lesion such as neoplasm or infection (Daffner and Pavlov 1992). Compressive forces occur along the concave margin of bone and may result in “compressive” fractures. Typical compression fracture sites are the inferior part of the femoral neck, the calcaneus, the other tarsal

14 bones, the inferior pubic ramus, the metaphyseal area of the tibia, and the proximal end of the fi rst metatarsal (Meurman 1981b). Tensile forces are produced along the convex side of bone and may result in “tension or distraction” fractures. Typical fracture sites are the superior cortex of the femoral neck, and the shaft of long bones (Boden and Speer 1997, Egol et al. 1998, Boden and Osbahr 2000, Lassus et al. 2002). High risk bone stress injuries occur on the tensile side of bone, or in bones with vulnerable blood supply areas. They can present a risk of progression to complete fracture, further displacement, delayed union or non-union, bony fragmentation, and, later, development of osteoar- thritis. They require prompt diagnosis, aggressive treatment, and occasionally even internal fi xation. The high risk areas include the medial malleolus, the tarsal navicular, the base of the fi fth metatarsal bone, and the fi rst digit sesamoids (Boden and Osbahr 2000, Lassus et al. 2002). Low risk bone stress injuries can be diagnosed by a thorough history, physical examination, and, usually, radiography is adequate. With early diagnosis and treatment observing restriction of activity, the prognosis is favorable (Boden et al. 2001). The low risk areas (with few exceptions) include the sacrum, the pubic rami, the femoral shaft, the tibial shaft, the fi bula, the calcaneus, and the metatarsal shaft. Internal derangement of joints is a determinant used in those conditions that disturb the action and function of extraspinal articulations (Resnick and Kang 1997). Internal derangement of the knee has commonly been used to refer to the condition of a painful knee joint, for which the more specifi c, pathology-oriented diagnoses are meniscal tear, ligament rupture, synovial lesion, osteochondral lesion, or articular cartilage degeneration. Cancellous bone (spongy or trabecular) consists of a meshwork of primary (longitudinal) and secondary (transverse) trabeculae separated by fatty or hematopoietic marrow. Typically found at the ends of long bones and inside cortical bone, cancellous bone consists of internal rods or plates forming a three-dimensional branching lattice. High surface- to-volume ratio makes a bone metabolism control possible. The cells lie primarily between lamellae or on the surfaces of the trabeculae, and are directly infl uenced by adjacent bone-marrow cells (Buckwalter et al. 1996).

15 Cortical bone (dense or compact) has a solid, compact architecture interrupted only by the narrow canals of the Haversian systems contain- ing neurovascular bundles. It typically presents along the outer casing of long bones and has a low surface-to-volume ratio. The cells are completely surrounded by bone matrix. Cortical bone comprises two types of surfaces; endosteum on the inner side faces the bone marrow and the trabeculae, and periosteum on the outer side faces the surrounding soft tissues (Buckwalter et al. 1996). Risk factor (exposing factor) is a determinant of disease development, an attribute or agent suspected to be related to the occurrence of a particular disease. Identifi cation of risk factors can result in a better understanding of the pathways leading to disease acquisition and crea- tion of preventive strategies. Person-time (PT or exposure time) is the number of new cases in defi ned population during defi ned time. Risk (probability of event occurrence) is the probability that an event (e.g. development of disease) will occur per exposure time. Incidence rate (IR) is the rapidity with which new cases of a particular disease arise within a given population. Population at risk comprise persons who are susceptible to a particular disease but who are not yet affected.

2.3. Bone marrow, epiphyses, metaphyses

Cancellous bone consists of a meshwork of primary (longitudinal) and secondary (transverse) trabeculae separated by fatty or hematopoietic marrow (Figure 1). The interface between the marrow and the trabecular bone is very active metabolically (Resnick and Kang 1997). Two forms of bone marrow exist; hematopoietically active red marrow and hematopoietically inactive yellow marrow. Red marrow consists of 40% water, 40% fat, and 20% protein. Yellow marrow consists of 15% water, 80% fat, 5% protein (Vogler and Murphy 1988). The amount of red marrow versus yellow marrow depends on the age of the person, the site of bone, and the health of the individual. At birth, red marrow is present throughout the skeleton. With increasing age and during growth, the normal conversion process of red marrow to fatty marrow takes place gradually, diminishing the amount of red marrow. Marrow conversion in the diaphyses occurs over the ﬁ rst decade of life, and, in the distal meta-

16 physes, over the second decade of life. The replacement occurs from distal bones to central bones, and from distal end of bone to the proximal end, or from the center of the shaft to both directions (Resnick and Kang 1997). The marrow conversion is usually complete by the age of 20 to 25. In the adult type marrow, red marrow is present in the metaphyses of the femora and the humeri, in portions of the vertebrae, ribs, sternum, clavicles, scapulae, scull, and innominate bones. Around the knee, the amount of hematopoietic marrow is greater in the distal portion of the femur than in the proximal portion of the tibia. The MR imaging appearance of bone marrow depends on the relative proportions of hematopoietic and fatty marrow (Resnick and Kang 1997).

Figure 1. The major features of a typical long bone. 1. head, 2. neck, 3. shaft (diaphysis), 4. periosteum, 5. articular cartilage, 6. epiphyseal line, 7. spongy bone, 8. compact bone, 9. endosteum and 10. medullary cavity. (© Research center of military medicine / Larni HM 2006)

17 2.4. Bone reaction to stress

Understanding the anatomy and physiology of bone and bone marrow is essential to the interpretation of imaging studies in patients with skeletal disorders (Resnick and Kang 1997). Bone is a dynamic tissue; once formed or modeled it is never metabolically at rest but is continually remodeling. In this lifelong renewal process, osteoclasts remove bone and osteoblasts replace bone without signifi cantly affecting the bone shape or density (Frost 1990a, Frost 1990b). Remodeling occurs on the periosteal, Haversian, cortical-endosteal, and trabecular bone surfaces (Figure 2). Bone undergoes remodeling in response to stress as specifi ed in Wolff´s law (Wolff 1892). As new stress is placed upon a bone, the bone tissue attempts to remodel to adapt to the new stress; it reacts to stress at fi rst by early osteoclastic activity followed by strengthening osteoblastic activity in which weaker or injured bone is replaced with stronger new bone (Johnson et al. 1963, Jones and Harris 1989). The mechanism of this phenomenon is not completely understood, but, on a microscopic level, it appears to be stimulated by microfractures (Ma- naster et al. 2002). If physical stress continues during this “osteoclast dominant phase,” the microfractures can coalesce into a discrete fracture. The main functions of remodeling are to adapt bone to mechanical loading, to prevent accumulation of microfractures or fatigue damage, and to maintain constant blood calcium levels (Frost 1989a, Frost 1989b, Frost 1991). According to Boden and colleagues (Boden and Speer 1997, Boden and Osbahr 2000) bone stress injuries represent an imbalance of bone resorption over bone formation. Histologically, during the fi rst week, stress fractures usually show active osteoclasts, an osteoclastic resorption without microfracture. Dur- ing the second week, periosteal callus and occasional endosteal callus formation follow. It has been stated that if the activity that caused the initial problem ceases, then no fracture line can be seen, but if the stress continues, a fracture of the cortex appears by the end of the second week in nearly one third of the cases (Johnson et al. 1994). Usually, resorption is complete at the end of three weeks and callus formation continues from the second week to its maximum point at six weeks (Johnson et al. 1963). The marrow edema seen on MR images may represent this accelerated remodeling or stress reaction, or bone stress injury, and at the histologic examination, possibly interstitial fl uid, microfractures and/ or blood (Vogler and Murphy 1988, Fazzalari 1993, Li et al. 1998). The

18 low signal line seen on MR images can represent either a stress fracture that is not seen on plain radiographs, or a complete fracture.

Figure 2. Diagram of the arrangement of the mature cortical bone. 1. Haversian system, 2. lacunae, 3. canaliculi, 4. cementing line, 5. inner circumferential lamellae, 6. outer circumferential lamellae, 7. interstitial lamellae, 8. blood vessel, 9. endosteum, 10. Haversian canal, 11. Volkmann’s canal, 12. ﬁ - brous and 13. cambium layer of periosteum. (© Research center of military medicine / Larni HM 2006)

19 2.5. Incidence of bone stress injuries

The reports on the general incidence of all lower extremity bone stress injuries has varied from 2% to as high as 31% (Sahi 1984, Milgrom et al. 1985, Milgrom et al. 1986, Jones et al. 1989, Bennell et al. 1996, Bennel and Brukner 1997). In the Finnish Defence Forces, the incidence of bone stress injuries has varied from 1.0% to 24.5% depending on the military branch and training program (Sahi et al. 1996). In the U.S. armed forces, the incidence of lower extremity stress fractures is 0.2 to 4.0% for male recruits and from 1 to 7% for female recruits (Greaney et al. 1983, Jones et al. 1993, Beck et al. 1996, Almeida et al. 1999, Shaffer et al. 1999, Kelly et al. 2000). Reports of the Israeli military population show remarkably higher bone stress injury rates, ranging from one-third to nearly one- half of the recruits (Milgrom et al. 1985, Giladi et al. 1991, Milgrom et al. 1994). About 500-600 sportsmen are affected by a bone stress injury annually in Finland. In the general athletic population, occurrence of fatigue bone stress injury is less than 3.7% (Hulkko and Orava 1987). In a typical sports medicine practice, bone stress injuries account for 10% to 20% of all injuries (Orava 1980, McBryde 1985, Jones et al. 1989). Variations in the incidence and anatomic location are due to differences in military branches, training programs, sports, length and intensity of activities, and identiﬁ cation methods (active/passive). Historically, the varying terminology has been interchangeable; the main question being at what stage of the physiologic reaction to physical stress should a patient be considered as having a stress fracture.

2.6. Etiology of bone stress injuries

Bone stress injuries are recognized to occur in many anatomic locations, predictable from the types of activities which may produce them (Daff- ner 1978). Fewer cases of stress fractures have been described in army recruits and athletes who have been in good physical condition, and who have participated in sports before their military service, or in heavy physical exercise, like marathon running (Gilbert and Johnson 1966, Meurman 1980a, Greaney et al. 1983, Hohmann et al. 2004). In a prospective study of 839 scintigraphic abnormalities in U.S. Marine recruits

20 by Greaney et al. (Greaney et al. 1983), recruits with a prior history of long-distance running (1 mile/day, >4 days/week) before military training developed signiﬁ cantly fewer stress fractures than did non-runners. In MRI study of eight marathon runners by Hohmann et al. (>3000 km annually during the 5 years preceding the baseline MRI) (Hohmann et al. 2004), there were no signs of increased internal stresses on bones and joints after long-distance running that could be detected on MR images. In spite of the wide research, no single cause for bone stress injuries or stress fractures can be given, whether they occur predominantly owing to increased impact stresses and fatigue of supporting structures, or to contractile muscular forces acting on and deforming the bone (Math- eson et al. 1987, Krivickas 1997). The etiological factors have been divided into extrinsic and intrinsic. Extrinsic causes include training errors, such as excessive volume, excessive intensity, duration and change of each strain cycle, excessive muscle fatigue, inadequate recovery, faulty technique, and impact attenuation, like training surfaces and training conditions, footwear and other equipment (Burr and Milgrom 2001). Intrinsic causes include muscle fatigue leading to the transmission of excessive forces to the underlying bone (Blickenstaff and Morris 1966, Yasuda et al. 1992, Boden and Osbahr 2000), muscle imbalance, lack of ﬂ exibility due to generalized muscle tightness, focal areas of muscle thickening or restricted joint range of motion, decreased bone strength due to low bone mineral density (Carter et al. 1981), and psychological factors, such as excessive training, nutritional intake, or eating disorders (Matheson 1987, Bennell and Brukner 1997). Female recruits have been reported to exhibit more frequent involvement of pelvic bone stress injuries than male recruits (Kiuru et al. 2001). The bigger risk for females, as compared with males, to develop bone stress injuries has been explained by anatomic / biomechanic (wide pelvis, coxa vara, genu valgum), hormonal, and nutritional factors. The female athletic triad with disordered eating, amenorrhea, and altered bone mineral density is well known (Pentecost et al. 1964, Provost and Morris 1969, Daffner 1978, Sahi 1984, Matheson et al. 1987).

21 2.7. Clinical diagnosis of bone stress injuries

The key to diagnosis is a strong suspicion based on the patient´s history and clinical fi ndings (Gilbert and Johnson 1966). The classic history of a stress fracture is that of pain appearing in a particular activity, amelio- rated by rest and deteriorated when the activity is continued (Daffner 1978). The type and severity of symptoms are related to a specifi c bone, but may be confusing when the hip, the pelvis, the femur, or the navicular bone is involved. The clinical diagnosis of a bone stress injury is diffi cult if symptoms are insidious and pain is diffuse or radiating (Provost and Morris 1969, Sallis and Jones 1991, Shin et al. 1996). In a report by Salminen and colleagues (Salminen et al. 2003), fi ve displaced fractures of the middle or distal third of the femoral diaphysis involved pain radiating to the knee. The part of the femur that is surrounded by strong muscles is diffi cult to palpate, whereas the femoral and tibial condyles, the anterior tibial shaft, and the ankle and foot bones are easily palpable and swelling, and, generally, a sharply localized pain with surrounding tenderness can be found. To increase the accuracy of diagnosis of a femoral bone stress injury, a fulcrum test has been recommended (Johnson et al. 1994). This test may also be useful in assessing the healing response. With the patient seated on the edge of the examining table, gentle pressure is ap- plied to the dorsum of the knee with the clinician’s hand. The opposite arm is used as a fulcrum under the patient’s thigh and is moved from distal to proximal. When the arm is under the site of the stress fracture, the patient experiences pain and apprehension. Discomfort in the area of the fracture together with restriction of hip, knee, and ankle joint movements can be clinically identifi ed when the fracture occurs near a joint. In addition, there may be swelling and effusion of the adjacent joint (Morris and Blickenstaff 1967, Kuusela 1980). Among recruits, the reaction to stress pain depends on the individual’s motivation for military service (Hallel et al. 1976), the ambitious may have higher pain thresholds. There may be a considerable interval between the onset of pain and the request for medical consultation, which can be further prolonged by the lack of a specifi c trauma-history. The clinical fi ndings as well as imaging fi ndings may vary depending on the phase of the pathophysiological spectrum in the bone stress injury (Anderson and Greenspan 1996). The fact that bilateral, multiple, and asymptomatic high grade bone stress injuries occur makes the clinical

22 diagnosis even more challenging (Meurman and Elfving 1980b, Kiuru et al. 2001). In a study of Meurman and Elfving, 40% of the recruits had multifocal stress fractures (Meurman and Elfving 1980a).

2.8. Diagnostic imaging of bone stress injuries

2.8.1 Radiography

Radiographs are indicated for the ﬁ rst imaging examination whenever there is a suspicion of a stress fracture or overuse injury (Bergman and Fredericson 1999). Radiography has been used as the primary imaging tool, but because of its low sensitivity in the early stages (10%) and in the follow-up (30% to 70%) of bone stress injuries (Prather et al. 1977, Orava 1980, Greaney et al. 1983, Rupani et al. 1985, Matheson et al. 1987, Nielsen et al. 1991), diagnosis has often been based on bone scintigraphy (Daffner 1978, Anderson and Greenspan 1996). In radiographs, the appearance of a stress fracture in the cortical bone may vary from a gray cortex sign or callus to a true fracture line (Mulligan 1995). The appearance of a stress fracture in the cancellous bone involves a focal zone of sclerosis due to secondary peritrabecular callus with possibly a solid or thick laminal periosteal reaction during the healing phase (Savoca 1971, Daffner 1978, Daffner 1984, Mulligan 1995).

2.8.2 Scintigraphy

In bone scintigraphy, imaging is based upon the detection of radiation emitted from radiopharmaceuticals, most commonly 99m Tc-phosphate analogs, inside the patient. Zwas and his study group graded the scintigraphic areas of increased uptake by the 1 to 4 rating system in which the size of the foci was assumed to reﬂ ect the amount of microdamage (Zwas et al. 1987). Pathologic uptake may be detected as early as 6 to 72 hours after injury. In a study of 250 U.S. Marine recruits by Greaney et al. (Greaney et al. 1983), the radiographic conﬁ rmation of stress fractures seen in scintigraphs was dependent on the duration of symptoms, anatomic site of the foci, whether it was cortical or cancellous, and on the

23 intensity of the bone scan foci. When compared to MRI, previous studies have shown that the sensitivity of both bone scintigraphy and MRI are excellent, but the speciﬁ city of bone scintigraphy is signiﬁ cantly lower than that of MRI (Anderson and Greenspan 1996, Kiuru et al. 2002). Moreover, both x-ray and scintigraphy use ionizing radiation. In bone scintigraphy, the radiation dose is equal to two years of background radiation (Kanstrup 1997).

2.8.3 Magnetic Resonance Imaging

Magnetic resonance imaging is an important imaging modality for the musculoskeletal system, since various tissues display distinctive signal intensities on T1- and T2–weighted images. The signal intensities for yellow marrow and red marrow can be distinguished owing to their different water and fat contents. For detecting joints, osseous, and soft tissue injuries, and to detect and differentiate the various forms of marrow pathology, MR imaging is not only highly sensitive but also an extremely specifi c imaging modality compared to radiographs and scintigraphy (Anderson and Greenspan 1996, Kiuru et al. 2002). In bone stress injuries, fat suppression techniques, T2-weighted images with fat suppression and STIR images, are the most commonly used means for detecting increased amounts of extracellular water controlling the degree of signal intensity alteration (Vogler and Murphy 1988). On MR images, endosteal marrow edema may be a highly sensitive, early sign of a stress-related bone stress injury (Arendt and Griffi ths 1997). However, marrow edema is a non-specifi c fi nding that can also be associated with bone tumors, bone infection, bone bruising (bone marrow contusion), and even found after heavy physical exercise (Pathria and Isaacs 1992, Schweitzer and White 1996, Lazzarini et al. 1997, Lohman et al. 2001). An MRI fi nding specifi c to a bone stress injury is a low-signal fracture line (Lee and Yao 1988). With a stress fracture present, T1- and T2- weighted MR images will show a band or area of low signal intensity that extends from the involved cortex into the medullary canal, most likely representing either microfractures or bone sclerosis (Lee and Yao 1988). T2-weighted MR images with fat suppression and STIR images can show a high signal intensity intramedullary fi nding, either alone or in combination with periosteal edema, soft tissue edema, or around the low-signal fracture line. In T1-weighted images, this marrow edema is low in signal. Callus is apparent as an intermediate cortical signal inten-

24 sity (Fredericson et al. 1995, Kiuru et al. 2001). A number of MRI stress fracture grading scales have been published for assessment of the various MRI signs (Lee and Yao 1988, Fredericson et al. 1995, Kiuru et al. 2001).

2.9. Differential diagnosis of bone stress injuries

Many inﬂ ammatory, infectious, vascular, tumorous, and exertional pathologic conditions of soft tissues and bones of the lower extremities may clinically mimic bone stress injuries (Table 1) (McBryde 1985, Milgrom et al. 1986). Differential diagnosis of distal thigh pain is extensive, including quadriceps tendon pain, patellofemoral pain, bone stress injury of the femoral shaft, supracondylar femur, and femoral condyle (Milgrom et al. 1986, Glorioso et al. 2002, Salminen et al. 2003). The proximity of a bone stress injury to the knee joint can confuse the clinical diagnosis with a lesion within a joint (Engber 1977, Schweitzer 1996, Glorioso et al. 2002); medial knee pain can raise a suspicion of a rupture of the medial meniscus (Milgrom et al. 1986, Vossinakis and Tasker 2000), increased intra-articular joint effusion may associate with bone stress injuries (Kuu- sela 1980), and an insufﬁ ciency bone stress injury in the knee has been reported simultaneously with intra-articular derangement (Sokoloff et al. 2001). Iliotibial band syndrome manifests itself as lateral knee pain, which may be sharp or burning on the lateral aspect of the knee when iliotibial tract slides over the lateral femoral epicondyle. The iliotibial tract or band is formed as an extension to the tendinous portion of the tensor fascia lata muscle and a portion of the gluteus maximus blend together, and inserts distally on Gerdy´s tubercle at the lateral tibia (John- son et al. 1994, Bergman and Fredericson 1999). Many different radiographic changes can mimic bone stress injuries. They can be divided into benign (osteoid osteomas, chronic sclerosing osteomyelitis, osteomalacia) and malignant (osteosarcoma, Ewing tumor, metastases). According to Buckley et al. (Buckley et al. 1995), serial radiographs taken during a period of several weeks to look for evidence of progressive fracture healing or increased bony destruction secondary to neoplasm may be an adequate approach to the treatment of stress fracture. Osteogenic sarcoma is the most common primary malignant bone tumor in adolescents (Manaster et al. 2002). Most osteosarcomas (90%) are metaphyseal in origin, but can be diaphyseal (Manaster et

25 al. 2002). Although metaphyseal in origin, the tumor frequently crosses the physeal plate to involve the epiphysis as well (Manaster et al. 2002). In osteosarcoma, periosteal reaction is that of an aggressive lesion and demonstrates speculation, thin lamination, coarse deposition, and occasionally Codman triangles. In time, the lesions progress and do not resemble a healing fracture. In the case of bone tumors, there is usually no history of recent increase in physical activities, such as running or marching. A distal femoral cortical defect appears commonly at the insertion of the tendinous fi bers of the adductor magnus muscle, and the femoral cortical excavations at the osseous site of attachment of the medial head of the gastrocnemius muscle. These cortical irregularities with suggested stress-related pathogenesis occurring in active individuals may simulate neoplasm (Resnick and Greenway 1982, Hyman et al. 1995). In case of an unclear periosteal reaction or an irregular cortex, MRI is useful for excluding neoplasm within the medullary canal or the surrounding soft tissues in a differential diagnosis between a bone stress injury and a bone tumor (Lee and Yao 1988) In MRI, a thin rim of high signal intensity on T2-weighted images along the medial periosteum, involving the proximal to mid-femoral shaft alone or with the medullary or cortical signal, has received various interpretations. Some authors consider it to be the adductor avulsion syndrome (thigh splints) corresponding to tibial shin splints, whereas others believe it represents the early stage of a developing stress fracture (Ozburn and Nichols 1981, Anderson et al. 2001, Kiuru et al. 2002). In case of pain manifesting along the medial border of the tibia, the clinical diagnosis between bone stress injury, traction periostitis (shin splint syndrome), and chronic exertional compartment syndrome is diffi cult, because they can occur separately or together (Kiuru et al. 2003a). In MRI, it is suggested that the periostitis seen with shin splint syndrome represents an early injury in a spectrum that also includes bony stress reaction and, later, a stress fracture (Bergman and Fredricson 1999, Kiuru et al. 2003a). Around the ankle and foot area, in addition to bone stress injuries, MRI can be useful for identifying some other causes for painful syndromes such as osteochondral fractures, sinus tarsi syndrome, painful accessory navicular, osteonecrosis of the talus, navicular or metatarsal heads, posterior impingement syndrome, bursitis, accessory muscles, Morton´s neuroma, and plantar fi bromatosis (Steinbach 1998).

26 TABLE 1. Differential diagnosis of bone stress injuries.

Conditions that may mimic bone stress injuries.

Patellofemoral pain Quadriceps tendinosis Iliotibial band syndrome Bursitis Osteoid osteomas Chronic sclerosing osteomyelitis Osteomalacia Distal femoral cortical defect Femoral cortical excavations Bone tumors Osteosarcoma Ewing tumor Metastases Internal derangement of the knee Morton´s neuroma Sinus tarsi syndrome Osteochondral fractures Painful accessory navicular Posterior impingement syndrome Plantar ﬁ bromatosis Inﬂ ammatory diseases Infectious conditions Transient bone marrow edema Osteonecrosis Thigh splint (adductor avulsion syndrome, traction periostitis) Tibial shin splint

27 2.10. Asymptomatic bone stress injuries

Bone marrow edema seen on MRI can be present in asymptomatic physically active individuals without previous trauma (Lazzarini et al. 1997, Lohman et al. 2001). Biomechanical abnormalities have been thought to explain why some runners develop the asymptomatic alterations of the bone marrow seen on MRI (Schweitzer and White 1996). Depending on the study design, these asymptomatic grade I MRI lesions have been considered either as bone stress injuries (Kiuru et al. 2003b) or, in case of a positive ﬁ nding in MRI without any clinical correlation, as increased bone remodeling or stress reaction in response to increased stress and not as a bone stress injury. Bone marrow edema has been reported to occur on the asymptomatic side in patients with multiple bone stress injuries, but also fracture line and callus have been reported on the asymptomatic side (Kiuru et al. 2003b). In the prospective study of Milgrom et al., 69% of the femoral stress fractures but only 8% of the tibial stress fractures were asymptomatic (Milgrom et al. 1985). The high incidence of asymptomatic femoral bone stress injuries has been explained by the lower sensitivity of the femoral periosteum as compared with that of the tibia (Milgrom et al. 1985, Burr and Milgrom 2001), which, together with deeply situated nonpalpable femoral bone, causes difﬁ culties to delineate between muscle and bone pain. In case of multiple bone stress injuries, a painful bone stress injury can “mask” the symptoms from other sites. However, to avoid complications, early diagnosis of asymptomatic bone stress injuries, especially those in the high-risk area, is important. For accurate diagnosis of bone stress injuries and to ensure appropriate treatment, simultaneous examination of the entire pelvis and both proximal femurs by MRI has been recommended (Kiuru et al. 2003b).

2.11. Symptomatic bone stress injuries

Exercise-induced musculoskeletal pain is a typical reason to seek medical advice. Ideally, the symptoms and clinical ﬁ ndings derive from a site of a bone stress injury. However, all patients cannot recall a speciﬁ c activity that could be associated with the onset of symptoms, and the symptoms can be vague or in a different site than the bone stress injury. Conse-

28 quently, not only does the patient’s history play an important role, but also the knowledge about the nature of bone stress injuries and their typical signs and symptoms relating to different anatomic locations is valuable. The majority of bone stress injuries have been reported to develop during the basic training period in military service, when trainees deal with an abrupt increase of physical activity that they are not accustomed to in their civilian life (Morris and Blickenstaff 1967, Geslien et al. 1976, Engber 1977, Greaney et al. 1983, Sahi 1984, Milgrom et al. 1985, Mil- grom et al. 1986, Muralikuttan and Sankarart-Kutty 1999). Bone stress injuries are common in “weekend-athletes” who have recently started a new training system, or have increased their training intensity or duration. Unfortunately, the duration or intensity of pain cannot always predict the seriousness of a bone stress injury. In previous studies, for instance, duration of pain before diagnosis of undisplaced and displaced femoral bone stress injuries has been reported to vary from a few days to several months (Provost and Morris 1969, Giladi et al. 1985, Matheson et al. 1987, Visuri et al. 1988, Visuri and Hietaniemi 1992, Clement et al. 1993, Shin and Gillingham 1997, Salminen et al. 2003).

2.12. Bone stress injuries in the lower extremity

The lower extremities are the most common site of bone stress injuries (Table 2). The frequency and location of bone stress injuries differ in various materials of military trainees and athletes. The most common anatomic locations are the tibial shaft, metatarsal bones, calcaneus, pubic rami, and femoral neck (Devas 1965, Orava et al. 1978, Kuusela 1980, Orava 1980, Meurman 1981a, Sahi 1984, Fullerton and Snowdy 1988, Bennell et al. 1996, Korpelainen et al. 2001, Kiuru et al. 2003b). In a scintigraphic study of 320 athletes by Matheson et al. (Matheson et al. 1987), the tibia (49.1%), the tarsals (25.3%), and the metatarsals (8.8%) were the most commonly involved bones. Of all bone stress injuries in athletes, the femur was the fourth most common site of stress fractures (Matheson et al. 1987).

29 2.12.1 Bone stress injuries of the hip

In 1965, Devas described two types of femoral neck stress fractures; compression fractures and distraction (tension) fractures (Devas 1965). Ac- cording to Devas, stress fractures of the femoral neck that appear on the inferior aspect (compression fractures) may propagate across the femoral neck at an angle that is not perpendicular to the axis of the femoral neck. These compression fractures are thought to be mechanically stable or presenting a lower risk for displacement, and are more common in young patients. Fractures that appear on the superior aspect (distraction fractures, tension fractures) present an increased risk for displacement with continued stress (Egol et al. 1998, Boden and Osbahr 2000, Lassus et al. 2002). The cortical crack of the tension side may advance across the femoral neck as a fracture line perpendicular to the axis of the femoral neck. Tension fractures are more common in patients over 60 years of age (Boden and Speer 1997). The typical symptom is hip pain or pain in the proximal part of the leg. Displacement of the femoral neck can occur even without preceding symptoms (Luchini et al. 1983). A dislocated stress fracture in a young person may have disastrous consequences, from a longlasting absence from physical activity to osteonecrosis of the femoral head, delayed union, malunion, and osteoarthrosis (Meurman 1981a, Visuri et al. 1988, Boden and Speer 1997, Visuri 1997, Egol et al. 1998, Salminen et al. 2003). In prior studies among military recruits, the proportion of femoral neck stress fractures of all diagnosed stress fractures has been under 10 % (Erne and Burckhardt 1980, Fullerton and Snowdy 1988, Volpin et al. 1990, Egol et al. 1998). The underlying pathomechanisms of femoral neck bone stress injuries are unclear. Malalignment of the lower extremity, i.e. varus alignment of the hip producing a focal concentration of mechanical stress in the femoral neck and alterations in the normal hip biomechanics, are potential risk factors for femoral neck bone stress injuries (Carpintero P et al. 2003, Hohmann E et al. 2004).

2.12.2. Bone stress injuries of the femoral shaft

In earlier studies based on plain radiographs and bone scans, 15% to 25% of all bone stress injuries in military recruits (Giladi et al. 1985, Volpin et al. 1989) and an estimated 10% of the stress fractures in athletes occur in the femur (McBryde 1975). The proportion of femoral shaft fatigue

30 fractures of all stress fractures, undisplaced and displaced fractures in- clusive, varies between 3% to 43% in military recruits (Hallel et al. 1976, Meurman et al. 1981, Schmidt-Brudvig et al. 1983, Giladi et al. 1985) and between 3.5% to 21% in athletes (Orava 1980, McBryde 1985, Johnson et al. 1994). In the reports of older military studies, the middle and distal shafts have emerged as the most commonly affected anatomic locations (Provost and Morris 1969, Hallel et al. 1976). In contrast, in previous studies among athletes, the proximal third has been reported to be the most commonly affected anatomic location (Hershman et al. 1990, Clem- ent et al. 1993, Johnson et al. 1994). Prior studies on exercise-induced femoral bone stress injuries suggest an even distribution of bone stress injuries between males and females (Luchini et al. 1983, Loyd-Smith et al. 1985). According to Boden and Speer (Boden and Speer 1997), femoral shaft bone stress injuries are thought to be related to training errors. The junction of the proximal and middle thirds of the femoral shaft (mid- medial or posteromedial cortex of the proximal femur) is the insertion of the adductor muscles, which are probably causative factors for bone stress injuries at this site if subject to repetitive avulsive stress (Meurman et al. 1981c, Boden et al. 2001). Since most of them can be treated conservatively, femoral shaft stress fractures in general are categorized as one of the low risk stress fractures (Boden et al. 2001). Some studies, however, have shown that displaced femoral shaft fractures may predict adverse consequences and prolonged morbidity as well (Salminen et al. 2003). That the femur is most anteriorly curved in the distal third where it is susceptible to anatomic and dynamic stresses, distal femoral stress fractures have the potential to displace (Provost and Morris 1969).

2.12.3. Bone stress injuries of the knee

The relationship between knee pain and stress fractures has been well established (Geslien GE, 1976, Hensal et al. 1983, Milgrom et al. 1986, Weber et al. 1988, Boden et al. 1997, Glorioso et al. 2002). Supracondylar bone stress injuries typically occur in the posteromedial cortex (Schmidt- Brudvig et al. 1983, Yasuda et al. 1992, Muralikuttan and Sankarart- Kutty 1999, Glorioso et al. 2002). The medial cortex is the tension side of the femoral supracondylar region in most individuals who have mild genu valgum (Glorioso et al. 2002). In a valgus alignment, the adductor

31 muscles and the medial collateral ligament may create opposing tensile forces that initiate the stress fracture. As with more common stress fractures, an abrupt increase in the intensity or duration of training is often reported. The reported cases of femoral condylar bone stress injuries in young adults are rare, whereas involvement of the medial tibial condyle is better known (Table 2). The proximal medial condyle of the tibia is the major weight-bearing portion of the bone (Greaney et al. 1983). The ﬁ rst case of patellar stress fracture in soldiers was described by Muller in 1943, and in athletes by Devas in 1960 (Muller 1943, Devas 1960). Exercise-induced stress fractures of the patella can be classiﬁ ed into longitudinal and transverse fractures, the latter constituting the majority of cases in the literature (Mason et al. 1996).

2.12.4. Bone stress injuries of the ankle and foot

Distal tibial bone stress injuries typically occur adjacent to or around the former epiphyseal line. Zlatkin et colleagues reported complicating stress fractures (four in the distal tibia and two in the calcaneus) occurring distal to the site of a healing traumatic fracture of the tibia or fi bula. All fi ve patients (age range 5 to 38 years) had disuse osteopenia at the time of appearance of the stress fractures, and they had recently begun partial or full weight-bearing (Zlatkin et al. 1987). Radiographically, stress fractures of the distal tibia appeared as lin- early oriented, dense bands of sclerosis, and they all healed without complications. Malalignment of the ankle has been reported to present in stress fractures of the distal tibia, distal fi bula (valgus deformity of the ankle), lateral metatarsal bones (varus alignment of the ankle), and medial metatarsal bones (valgus alignment of the ankle) in patients with rheumatic infl ammatory arthritides (Mäenpää et al. 2002). Distal fi bular bone stress injuries are considered common in athletes with a cavus- type foot, and in those who train on hard surfaces (Burrows 1948, Devas and Sweetnam 1956, McBryde 1975). The majority of distal fi bular bone stress injuries occur few centimeters proximal to the syndesmosis (Bur- rows 1948). Any bone of the foot may be involved with bone stress injury, even in unusual areas such as the cuboid, and the great toe. Calcaneal fatigue bone stress injuries have been encountered in activities such as parachute jumping or prolonged standing, and they are reportedly common

32 in military recruits (Greaney et al. 1983, Meurman 1981a). Navicular bone stress injuries occur typically in high-performance athletes. The central third of the tarsal navicular being relatively avascular when compared to the remainder of the bone may play a role in the development of a stress fracture (Sanders et al. 2004).

TABLE 2. Previous studies about bone stress injuries of the femoral neck, femoral shaft, knee, and ankle and foot.

Author and year Location of No of Method No of Bilat- bone stress pa- BSI eral injury tients FEMORAL NECK Blickenstaff LD and Morris femoral neck 36 - 41 5 JM, 1966 Erne P and Burckhardt A, femoral neck 5 retrospec- 25 5 1980 tive Fullerton LR and Snowdy HA, femoral neck 49 prospective 54 5 1988 Visuri T et al. 1988 femoral neck 12 retrospec- 12 - tive Clement DB et al. 1993 femoral neck x / 71 prospective 8 / 74 -

FEMORAL SHAFT Provost RA and Morris JM, femoral shaft 35 38 3 1969 Volpin C and Petronius G, femoral shaft 8 prospective 11 3 1989 Visuri T and Hietaniemi K, femoral shaft 3 retrospec- 3- 1992 tive Clement DB et al. 1993 femoral shaft x / 71 prospective 39 / 74 - Boden BP and Speer KP, 1997 femoral shaft 1 retrospec- 1- tive Salminen ST et al. 2003 femoral shaft 10 retrospec- 10 - tive Kiuru MJ et al. 2003 proximal femoral 34 retrospec- 34 - shaft tive

KNEE Geslien GE et al. 1976 tibial condyle NA prospective 73 51 Engber WD, 1977 medial tibial 36 retrospec- 57 21 condyle tive

33 Hensal F et al. 1983 patella 1 retrospec- 21 tive Milgrom C et al. 1986 medial femoral 3 retrospec- 3- condyle tive Weber PC, 1988 femoral metaphy- 1 retrospec- 1- sis-epiphyse tive Jerosch JG et al. 1989 patella 1 retrospec- 1- tive Yasuda T, 1992 supracondylar 1 retrospec- 1- femur tive Teitz CC and Harrington RM, patella 2 retrospec- 2- 1992 tive Orava S et al. 1996 patella 5 retrospec- 5- tive Mason RW et al. 1996 patella 3 retrospec- 3- tive Boden BP and Speer KP, 1997 femoral condyle 1 retrospec- 1- tive Muralikuttan KP and Sanka- supracondylar 2 retrospec- 2- rart-Kutty M, 1999 femur tive Vossinakis IC and Tasker TPB, medial tibial 1 retrospec- 1- 2000 condyle tive Mayers LB et al. 2001 patella 1 retrospec- 1- tive Glorioso JE et al. 2002 supracondylar 2 retrospec- 2- femur tive

ANKLE AND FOOT Orava S, 1980 metatarsals x / 185 prospective 36 - Meurman K, 1981a metatarsals 135 - 140 5 Matheson GO et al. 1987 metatarsals x / 320 retrospec- 33 - tive Matheson GO et al. 1987 tarsal bones x / 320 retrospec- 94 - tive Volpin C and Petronius G, metatarsals x prospective 10 - 1989 Volpin C and Petronius G, tarsal bones x prospective 3 - 1989 Bennell K et al. 1996 navicular bone 4 / 20 prospective 4 -

34 3. Aims of the present study

1. To evaluate by clinical and MRI follow-up the occurrence of asymptomatic bone stress injuries, their clinical signiﬁ cance, and whether they all progress to stress fractures in elite-unit military recruits undergoing intensive physical training (I).

2. Based on MR imaging, to assess the incidence, location, nature and patterns of bone stress injuries of the knee in Finnish conscripts (II).

3. Based on MR imaging, to assess the incidence, location, nature and patterns of bone stress injuries of the femoral bone in Finnish conscripts (III).

4. Based on MR imaging, to assess the incidence, location, nature and patterns of bone stress injuries of the ankle and foot in Finnish conscripts (IV).

35 4. Materials and methods

The ﬁ rst prospective cohort study (I) was conducted at the Utti Jaeger Regiment, and at the Department of Radiology and the Research Insti- tute of Military Medicine, Central Military Hospital Tilkka, Helsinki, Fin- land. Three retrospective studies (II-IV) were carried out at the Depart- ments of Radiology and Surgery and the Research Institute of Military Medicine, Central Military Hospital Tilkka, Helsinki, Finland. All study designs (I-IV) were approved by the Defence Staff of the Finnish Defence Forces. Furthermore, the Study I was approved by the Ethical Commit- tee of Kymenlaakso Hospital District, and informed consent from each recruit for this study was received prior to participation.

4.1. Patients

Over the study period, between March 1997 and April 2004, all patients concerned were performing their military training as conscripts of the Finnish Defence Forces. All Finnish men become liable for a 6-, 9-, or 12- month-long military service at the age of 18. For women, military service is voluntary. Entering service takes place either in January or in July. Each year, on the average 26,500 male conscripts and 500 female conscripts undergo training in the Finnish Defence Forces. All conscripts receive the healthcare services of the military hospitals, and even if ﬁ rst treated at a civilian hospital, they are later transferred to a military hospital. The service area of the Central Military Hospital in Helsinki covers 75% of the conscripts in Finland (Figure 3).

36 Figure 3. The service area of the Central Military Hospital in Helsinki covers 75% of the conscripts in Finland.

The prospective study I focused on a group of military recruits from the male elite-unit of Finnish parachutists. From a larger pool (N=43) of volunteers from the Parachute School of the Utti Jaeger Regiment, 21 randomly chosen 19-22-year-old recruits took part in the study during their intensive training period in 2003. About 500 to 600 young men ap- ply each year to join the elite unit, of these 100 are chosen on the basis of physical and psychological tests. During the School’s ﬁ ve-month physical training period, mean amount of physical training is approximately 25 hours per week (15 hours heavy training, 10 hours lighter training), gradually intensifying toward the end of the period. Three physical ex-

37 aminations were performed for all patients of the study I by a garrison physician of the Parachute School of the Utti Jaeger Regiment, the fi rst before the intensive training period started, the second at 6 weeks from the start, and the third on completion of the 5-month training program. After each examination, the patients were referred to the Central Mili- tary Hospital for imaging studies. For the retrospective studies II-IV, the participants represented the general conscript population during their entire military service. All patients of the studies II-IV underwent a physical examination by a garrison physician at the primary military healthcare units, where plain radiographs were taken as well. Radiographic fi ndings for all patients turned out negative, yet, because of prolonged pain and unclarity in clinical diagnosis, they were referred to the Central Military Hospital for an orthopaedic consultation. At the Central Military Hospital, all the patients were physically examined by an orthopaedic surgeon, as well as MR imaged on a clinical basis. For the retrospective study II, a computer search was utilized to iden- tify patients meeting the criteria of exercise-induced knee pain causing disturbance in their military training, and leading to MR imaging of the knee between March 1997 through December 2002. Altogether, 1330 patients (34 females, 1296 males, age range 17 to 29 years; mean age 20 years) were identifi ed, with 1577 MR imaged painful knees. All cases (images and patient data) with bone stress injury fi ndings were retro- spectively re-evaluated concerning bone stress injury locations and bone stress injury types. Of the 1330 patients, 88 (1 female, 87 males; age range 19 to 29 years, mean age 21 years) exhibited bone stress injuries in 110 knees. For the retrospective study III, a computer search identifi ed 1857 consecutive patients who, after examination by an orthopaedic surgeon, were referred to a subsequent MRI examination of the pelvis, hips, femurs, and/or knees during 70 months, beginning in 1997. All 1857 patients were MR imaged on account of exercise-induced pain in the femur during the military service. Such pain was reported in the area of the hip, groin, thigh, or knee. Retrospective re-evaluations of all the cases (images and patient data) with bone stress injury fi ndings were performed with respect to the location and the type of injury. Based on MR images, 170 (23 females, 147 males; age range 18 to 29 years, mean age 20.4 years) out of 1857 patients were ascertained with a fatigue femoral bone stress injury.

38 For the retrospective study IV, all consecutive conscripts who, over a period of 86 months beginning in 1997, were referred to an orthopaedic surgeon for consultation and to a subsequent MRI examination concerning the ankle and/or foot were identiﬁ ed via computer search. In all patients, the reason for MR-imaging was exercise-induced pain in the ankle and/or foot during the military service. The cases (images and patient data) with bone stress injury ﬁ ndings were chosen for retrospective re-evaluation of the location and type of injury. Based on MRI, 131 out of a total of 267 patients (9 female, 122 male, age range 17 to 27 years, mean age 20 years) were established to have between one and eleven bone stress injuries per one foot in altogether 142 ankles and/or feet.

4.2. Clinical diagnosis

In all studies (I-IV), the physical examinations by the orthopaedic surgeon involved careful history taking and palpation. Joint movements and ligamentous stability of the lower extremities were noted, distal pulses and sensation were tested, and skin changes were recorded, as well as any aberrant ﬁ ndings. Moreover, in studies I-III, range of movement of the hip joints (I,III); range of movement (extension-ﬂ exion) and ligamentous stability (anterior-posterior and side stability) of both knee joints (I,II); range of movement of the ankle joints and ligamentous stability of the ankle (IV), as well as ability to walk around and jump on one foot (I-III) were evaluated.

4.3. Imaging methods

In the study I, all imaging took place at the Central Military Hospital, whereas in the studies II-IV, the radiographic examinations were conducted at primary military healthcare units before patients’ referral to the Central Military Hospital. The majority of the radiographs were con- ventional anteroposterior and lateral radiographs. Depending on the case, the following standard primary radiographs were obtained: Knee AP and lateral projections (I,II), pelvis and hip AP projection (I,III), femoral shaft AP and lateral projections (I,III), ankle AP (ankle 15-20º inversion) and lateral projections (I,IV), foot AP and oblique projections (I,IV).

39 In case of a suspected calcaneal bone stress injury, both AP and lateral projections were obtained (I,IV). MR imaging for all studies (I-IV) was carried out with a 1.0 T MR unit (Signa Horizon, GE Medical Systems, Milwaukee, USA) at the Central Military Hospital, Helsinki, Finland. Coils employed for scanning consisted of a body coil (pelvis and leg), a knee coil (knee and ankle), and a surface coil (foot).

4.3.1 Pelvis and hip

In the pelvic and femoral area, routine coronal T1-weighted spin-echo (SE) sequence images (repetition time (TR)/echo time (TE) = 600 msec/19 msec, with two signals averaged and a 256 x 224 matrix) were taken, followed by coronal and axial T2-weighted fast spin echo (FSE) sequences with fat suppression (3000-6200/75-80) (with two signals averaged and a 512-256 x 224 matrix). A coronal STIR sequence was also used (400/17, inversion time (TI) = 140 msec, with two signal averaged and a 256 x 224 matrix). The ﬁ eld of view (FOV) was 32 to 48 cm x 24 to 48 cm, and the slice thickness was 4.0 to 5.0 mm, with a 0.5 to 1.0 mm intersection gap.

4.3.2 Knee

A knee coil with a ﬁ eld of view (FOV) of 10-16 cm was used. Slice thickness was (3)-4 mm, with a 0.5 or 1.0 mm intersection gap. Sagittal proton density (PD) spin-echo (SE) sequence images with fat suppression (repetition time (TR)/echo time (TE) = 3400 msec / 17 msec, with two signals averaged and a 256 x 256 (516) matrix) or sagittal T1-weighted spin-echo (SE) sequence images (680/11, with two signals averaged and a 256 x 256 (512) matrix) were obtained. T2-weighted sequences with fat suppression were taken for the axial (2560/85, with two signals averaged and a 256 x 256 (512) matrix) and for the coronal images (4000-4600/72-90, with two signals averaged and a 256 x 256 (512) matrix).

4.3.3 Ankle and foot

The ankle and foot area was imaged on at least two different planes, most commonly sagittal and coronal T1-weighted spin-echo (SE) sequence images and T2-weighted fast spin echo (FSE) sequence images

40 with fat suppression, but additional sequences, such as STIR, were included as well. MR imaging of the ankle was performed on at least two different planes, of which the T1-weighted spin-echo (SE) sequence images (repetition time (TR)/echo time (TE = 500-680 msec/10-15 msce), with two signals averaged, and 256 x 192-224 matrix, and the T2-weighted fast spin echo (FSE) sequence images with fat suppression (TR = 4400-6000 msec/TE effective = 80-90 msec, with two signals averaged, echo train length 8-12, and 256 x 224 matrix) were the most common. The ﬁ eld of view (FOV) was 18-20 x 18-20 cm, and the slice thickness was 3.0-4.0 mm, with a 0.5 to 1.0 mm intersection gap. Also additional sequences, such as STIR, (5400 msec/17 msec, TI 140 msec, with two signals averaged, a 256 x 224 matrix, a FOV of 32-48 x 24-48 cm, a slice thickness of 4.0-5.0 mm, with a 0.5 to 1.0 mm intersection gap) were obtained.

4.4. Image interpretation

Evaluations of the MR images for all studies (I-IV) were performed by two radiologists (M.N., M.K.) in consensus, blinded to the clinical data, according to the bone stress injury location and the bone stress injury type. Bone stress injuries were graded as follows: Grade I endosteal marrow edema, Grade II periosteal edema and endosteal marrow edema, Grade III muscle edema, periosteal edema and endosteal marrow edema, Grade IV fracture line, and Grade V callus in cortical bone. Marrow edema was demonstrated as low or intermediate signal intensity on T1- weighted images and as high signal intensity on T2-weighted and STIR images, and a fracture line was determined by a low signal intensity line on all of these pulse sequences. Callus was demonstrated as an intermediate cortical signal intensity mass. Furthermore, in patients with bone stress injury of the neck, the neck-shaft angle was measured (III).

4.5. Statistical methods

Data analysis for all studies (I-IV) was performed using SPSS/Win (ver- sion 12.0, SPSS Inc, Chicago, Illinois, USA). The median time to orthopaedic consultation from the commencement of military training, and the median time to diagnosis of bone stress injury by MRI from entering

41 military service were calculated in study II. The median time to onset of ankle and/or foot pain from commencement of the military service was calculated in study IV. The median time from onset of pain to MRI was calculated in studies II, III, and IV. The results were presented as median values and deviations with minimum and maximum (II,IIl, IV). Crude bone stress injury rates were presented as patients with bone stress injuries per 100,000 person-years (calculated by dividing the number of patients with MRI diagnosed bone stress injuries per the number of person-days by the average training period expressed in 100,000 person- years). Person-days were calculated carefully by registering the dates of entry into and transfer or discharge from the Central Military Hospital of every conscript in the catchment area of the hospital. When calculating person-based incidences (I), the number of new persons with injuries in the population are identifi ed and counted (A = 88), and the net time, also referred to as person-time (PT) or exposure time, during which individuals are at risk for developing injuries is measured (PT = 85,318). I = A/PT x 100,000 = (88/85,318) x 100,000 = 103 / 100,000 person-years in military service. The total exposure time for the population at risk during the study period was 85,318 conscript-years in the studies II and III, and 104,340 conscript-years in the study IV. In Finland, every garrison reports their so called “medium strength” at the end of the month. The fi gure is obtained by adding up the number of daily allowances paid out and dividing the sum by the number of days in the month. Hence, each monthly fi gure represents the total person-time, not the total number of conscripts. For example, considering one garrison with 1000 conscripts, all serving for the full month, with no interruptions, no discharges and no newcomers, the exposure time totals 1000 person-months. Since the ac- tual number of conscripts cannot be determined, fi gures are based on daily levels converted into person-months, which equals the total person-time of exposure per year. For statistical analysis, various methods were implemented throughout the study. In study I, unpaired proportions test (t-test) was used to determine the difference between mean body mass index at the beginning of the intensive training period and at the end of the period (I). The limit for statistical signifi cance was set equal to 0.05. In study II, Mann-Whitney U-test was used to assess differences in the continuous skewed data between groups, while differences in the crosstables were determined using Fisher´s Exact test. The limit for signifi cance was set

42 equal to 0.05. In study III, differences in the continuous skewed data between groups were assessed by Kruskal-Wallis H-test, and differences in the crosstables were determined using the Pearson chi-square test. Correlation between the continuous skewed data and the ordinal data was tested using the Spearman´s correlation coeffi cient (Rs). The limit for signifi cance was set equal to 0.05. In study IV, associations between occurrences of bone stress injuries were tested with Pearson chi-square test or Fisher´s exact test, and Mann Whitney´s U-test was used to test differences in the skew continuous data between groups. The limit for statistical signifi cance was set equal to 0.05.

43 5. Results

5.1. The occurrence and clinical signiﬁ cance of asymptomatic bone stress injuries in elite recruits (I).

Examination of three MRI scans revealed a total of 75 bone stress injuries in 16 of the 21 recruits, with only 40% (30/75) of these injuries being symptomatic. The symptoms were associated not only with the location of the injury but also the MRI grade in terms that higher grades usually involved more symptoms compared to lower grades. Repeated clinical and MRI assessment indicated that 84% of asymptomatic grade I bone stress injuries healed (21/25), or remained as grade I and asymptomatic (3/25, 12%). Comparing the beginning and the end of the intensive training period, it was evident that bone stress injuries, symptomatic cases, and recruits with bone stress injury increased in numbers toward the end of the period.

5.2. The incidence, location, and nature of bone stress injuries in the knee (II)

Of the 1330 patients, 88 met the inclusion criteria, and 141 bone stress injuries were found in their 110 knees (Table 3, Figures 4-8). The incidence of bone stress injuries was 103 per 100,000 person-years. Twenty- two (25%) of the patients had bilateral bone stress injuries. In 25 (28%) of the patients, two solitary bone stress injuries in the same knee simultaneously were discovered, all cases involving the tibial plateau and the femoral condyles. The most common anatomic location for a bone stress injury was the medial tibial plateau (43/141, 31%), which was also the

44 most typical location for a fracture line (67%). A bone stress injury in the medial tibial plateau caused knee pain sooner after entering the military service than a bone stress injury elsewhere in the knee (p=0.014). There were no individual clinical ﬁ ndings indicating a particular relationship with bone stress injuries of the knee. Of the 1577 knees MR imaged, at least one bone stress injury was found in 7% (110/1577) of the cases, internal derangement in 8.6% (135/1577) of the cases, and in 84,3% (1329/1577) of the knees, no pathological or aberrant ﬁ ndings were found.

TABLE 3. Anatomic locations and MRI Grades (I-V) of bone stress injuries in the knee (With permission from AJSM sep 16, 2005).

Location MRI grade of bone stress injury

Gr I Gr II Gr III Gr IV Gr V Total % no of locations

Femur supracondylar 1 1 - 3 1 6 4% medial condyle 25 1 - 2 - 28 20% lateral condyle 1 - - - - 1 1% both medial and lateral condyle 11 - - - - 11 8%

Tibia medial condyle 14 2 1 26 - 43 31% lateral condyle 5 - -1-64% both medial and lateral condyle 31 - - 3 - 34 24% proximal shaft 1 - - 4 2 7 5%

Patella 3 2 - - - 5 4%

Total no of grades 92 6 1 39 3 141

% 65% 4% 1% 28% 2%

45 4% 8% 20%

24%

31%

Figure 4. Percentual distribution of the bone stress injuries in the knee (With permission from AJSM sep 16, 2005)..

Figure 5. 20-year-old male conscript suffering from left knee pain for 41 days. (A) Coronal T2 weighted fat saturated MR image reveals a high signal intensity indicating endosteal bone marrow edema both in the medial femoral condyle and in the medial tibial plateau (arrows). A tiny intraosseous fracture line is visible (arrowhead). (B) Axial T2 weighted fat saturated MR image demonstrates not only endosteal edema (arrow) but also periosteal edema (arrowhead). (With permission from AJSM sep 16, 2005).

46 Figure 6. 22-year-old male conscript suffering from right knee pain for 14 days. (A) Coronal T2 weighted fat saturated MR image demonstrates a fracture line (arrowheads) with associated bone marrow edema of the medial tibial plateau. (B) Sagittal T1 weighted image indicates a nearly horizontal, posteriorly wide, slightly irregular line of reduced signal intensity (arrowheads) in the proximal medial tibia around the epiphyseal line. (With permission from AJSM sep 16, 2005)

Figure 7. 20-year-old male conscript suffering from right knee pain for 14 days. (A) Coronal T2 weighted fat saturated MR image indicating a transverse supracondylar femoral fracture line (arrows) with associated bone marrow edema and soft-tissue edema. (With permission from AJSM sep 16, 2005)

47 Figure 8. 19-year-old male conscript suffering from left knee pain for 21 days. (A and B) Sagittal and coronal T2 weighted fat saturated MR images reveal a high signal intensity indicating endosteal bone marrow edema of patella (arrows). (C) Axial proton density image demonstrates not only endosteal edema and periosteal edema but also soft tissue edema (Gr III) (arrow). (With permission from AJSM sep 16, 2005)

5.3. The incidence, location, and nature of bone stress injuries in the femoral bone (III)

Based on MR images, 170 (23 female, 147 male; age range 18 to 29, mean 20.4) out of 1857 patients met the inclusion criteria and were conﬁ rmed with a femoral bone fatigue stress injury. The total number of femoral bone stress injuries detected in these patients on MRI was 185, including ﬁ fteen (9%) patients with bilateral bone stress injuries and four patients with femoral condylar bone stress injuries (Table 4). The patients com- plained of pain in the hip, groin, thigh, or knee region. The incidence of femoral bone stress injury was 199 per 100,000 person-years, indicating an annual average of 29 patients in a new batch of conscripts.

48 TABLE 4. Bone stress injuries in femur. Distribution and grades of MRI ﬁ ndings. (Reprocuded with permission and copyright © of the British Editorial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

No (%) Sex Duration of Bilat- Grades of inju- F / pain before eral ries M MRI(days) I II III IV V Head 1 0000 1 (0.5%) 0 0 / 1 --- 14 / Neck 36 16 5 36 0 93 (50%) 5 28 74 Trochanteric area ------Subtrochanteric area ------Proximal shaft 22 840034 (18%) 3 7 / 24 30 Middle shaft 3 0100 4 (2%) 0 1 / 3 23 Distal shaft 0 1000 1 (0.5%) 0 0 / 1 21 Entire shaft area 2 0000 2 (1%) 1 0 / 1 7 Supracondylar area 1 1 - 3 1 6 (3%) 1 0 / 5 25 Condylar area 41 1-2-44 (24%) 5 1 / 38 70 23 / Total 106 27 10 41 1 185 15 147 % 57155220.5

The most common site for a femoral fatigue injury was the femoral neck (93%) followed by the condylar area (24%), the proximal shaft (18%), the supracondylar area (3%), the middle shaft (2%), the entire shaft area (1%), the distal shaft (0.5%) and the head (0.5%) (Figures 9- 13). Neither pain on palpation nor a distinct restriction of hip and/or knee movements could be ascertained in the patients. The only biomechanical factor measured, the neck-shaft angle, was normal, i.e. 125°-135°, in all patients conﬁ rmed with femoral neck stress injury. The relative proportion of femoral neck bone stress injuries displayed in the female conscripts (14/23, 60.9%) was higher than that of the male conscripts (74/147, 50.3%) (p=0.347). Marrow edema was found in 57% of the cases. In 22% of the cases demonstrating a low-signal fracture line (Gr IV), injuries were most common in the neck and the proximal shaft. Examination of the 93 neck bone stress injuries revealed marrow edemas (Gr I) and fracture lines (Gr IV) occurring to an equal extent in 39% of the cases. Of the neck fractures (n=36), a tiny fracture or cortical crack appeared perpendicular to the inferior cortex of the neck in 21 (58%) cases, a fracture line covered a half or two-thirds of the thickness of the neck in 6 (17%) cases, a fracture line covered the entire thickness of the neck in 8 (22%) cases, and, in one (3%) case, the fracture line was subcapital.

49 Head

Neck

Proximal

Middle

Distal

Supracondylare area 3 % Condylar area 24 %

50 Figure 10. 20-year-old-male conscript suffering from right hip and groin pain for 16 days. A) Coronal STIR image demonstrates a low signal intensity fracture line (arrow) in the femoral neck with associated high-signal intensity endosteal and periosteal edema (Reprocuded with permission and copyright © of the British Edito- rial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

Figure 11. 19-year-old-male conscript suffering from knee pain for 30 days. A) Coronal proton density fat saturated MR image demonstrates a low-signal fracture line (arrow) just above the epiphyseal line with surrounding high signal intensity marrow edema in the medial femoral condyle. B) The corresponding case in the sagittal T1 weighted image. The fracture line (arrow) with surrounding marrow edema (Reprocuded with permission and copyright © of the British Editorial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

51 Figure 12. 19-year-old-male conscript suffering from right hip and groin pain for 30 days. A) Coronal STIR image reveals a high signal intensity endosteal edema (arrow) in the proximal femoral shaft (Reprocuded with permission and copyright © of the British Editorial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

52 Figure 13. 19-year-old-male conscript suffering from left thigh and knee pain for 21 days. A) Coronal STIR image demonstrates endosteal (arrow) and periosteal edema (arrowheads) in the distal femoral shaft. B) Axial T2 fat saturated MR image shows also both endosteal (arrow) and periosteal edema (arrowhead) (Reprocuded with permission and copyright © of the British Editorial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

The median time from the onset of pain to the diagnosis of a bone stress injury was 30 days (range 3 to 270 days). In most cases, hip and groin pain were related to neck and proximal shaft bone stress injuries, and knee pain to condylar and supracondylar bone stress injuries (Table 5). The median time from the onset of pain to the conﬁ rmation of the bone stress injury by MRI was longer in patients with a bone stress injury

53 in the condylar area (70 days, range 14 to 270) than in patients with a bone stress injury in the neck (28 days, range 4 to 210) or in the proximal shaft (30 days, range 14 to 195) (p<0.001). A correlation was evident between bone stress injury grade and duration of pain preceding diagnosis by MRI (Rs = -0.258, p=0.001, n=170), indicating that a higher bone stress injury grade involved a shorter duration of pain preceding diagnosis.

TABLE 5. Location of pain in relation to anatomical location of bone stress injury in the femur (Reprocuded with permission and copyright © of the British Edito- rial Society of Bone and Joint Surgery) 87: 1385-1390, 2005).

Location of bone Location of pain stress injury hip groin thight knee Head 1 - - - Neck 83 8 1 1 Trochanteric area ---- Subtrochanteric area ---- Proximal shaft 30 2 2 - Middle shaft - - 4 - Distal shaft - - 1 1 Entire shaft area - - 1 1 Supracondylar area - - 1 5 Condylar area - - - 44 Total 114 10 10 52

5.4. The incidence, location, and nature of bone stress injuries in the ankle and foot (IV)

According to MRI evidence, 131 out of 267 patients (9 female, 122 male, age range 17-27 years, mean age 20) displayed between one to eleven bone stress injuries of the ankle and foot, including eleven patients (8%) with bilateral injuries. The incidence was 126 / 100,000 person-years. In the 142 feet and/or ankles involved in these patients, a total of 378 bone stress injuries were found in the following anatomic regions: tarsal bones (57.7%), metatarsal bones (35.7%), upper part of the talocrural joint (3.4%), sesamoids and accessory ossicles (2.6%), and digits (0.6%) (Table 6, Figure 14). In 89 (63%) feet, two or more bones were affected, and in 53 (37%) feet, a single bone was affected.

54 TABLE 6. MRI-gradings of the fatigue bone stress injuries of the talocrural joint and foot bones.

Per Location Grade n cent I II III IV V Talocrural joint - inferior tibia 6 - - 2 - 8 2.1 - inferior ﬁ bula 3 - 1 1 - 5 1.3 Tarsal bones - talus 39 6 2 8 - 55 14.6 - calcaneus 16 3 - 14 - 33 8.7

- navicular bone 24 5 2 2 - 33 8.7 - cuboid bone 15 2 1 - - 18 4.8 - 1st cuneiform bone 30 3 2 1 - 36 9.5 - 2nd cuneiform bone 23 2 1 - - 26 6.9 - 3rd cuneiform bone 13 3 1 - - 17 4.5 Metatarsal bones - 1st metatarsal bone 19 3 3 3 1 29 7.7 - 2nd metatarsal bone 20 2 14 2 4 42 11.1 - 3rd metatarsal bone 19 4 9 - 5 37 9.8 - 4th metatarsal bone 13 2 1 - 1 17 4.5 - 5th metatarsal bone 7 2 1 - - 10 2.6 Digits - distal phalanx of 1st digit 1 - - - - 1 <1.0 - proximal phalanx of 5th digit 1 - - - - 1 <1.0 Sesamoids and accessory ossicles - 1st digit sesamoids 3 - 1 2 - 6 1.6 - os trigonum 3 - - - - 3 <1.0 - os tibiale externum 1 - - - - 1 <1.0 Total (n) 256 37 39 35 11 378 Per cent 68 10 10 9 3 100

55 Figure 14. Percentual distribution of 378 fatigue fractures of the talocrural joint and foot according to the anatomical regions. Leg bones: 1. tibia and 2. ﬁ bula. The seven bones of the tarsus: 3. talus, 4. calcaneus, 5. navicularis, 6. cuboideus and 7. - 9. 1st, 2nd and 3rd cuneiform bones. Sesamoids and accessory ossicles (n = 7) are not included in the ﬁ gure (c.f. Table).

A talus bone stress injury was conﬁ rmed in 55 (39%) cases and a calcaneal bone stress injury in 33 (23%) cases. A bone stress injury of the second metatarsal bone was ascertained in 42 (30%) cases, of the third metatarsal bone in 37 (26%) cases, and of the ﬁ rst metatarsal bone in 29 (20%) cases. Considering the distal tibial bone stress injuries, six of them were observed around the tibial epiphyseal line, and two in the medial malleolus. In 9% of the cases, a fracture line was evident, most commonly in the calcaneus and in the talus, and in nearly 68% of the cases, endosteal edema was noticed, involving all bones of the imaged area (Figures 15-19).

56 Figure 15. 20-year-old conscript suffering from left foot pain for 20 days. A) In the coronal view, marrow edema in proximal part of the second metatarsal bone (arrow) and adjacent soft-tissue edema (small arrow) are seen. Subtle marrow edema (arrowheads) is also visible on the other bones of the foot. B) The same case in the sagittal view.

57 Figure 16. 21-year-old male conscript suffering from bilateral ankle pain for 14 days. MRI revealed bilateral grade IV bone stress injury of the distal tibia. A) Sagittal T1 weighted spin echo image demonstrates a linear band of low signal intensity representing fracture line (arrow) above epiphyseal line (small arrow) of the right distal tibia posteriorly. The fracture is surrounded by poorly deﬁ ned zone of low signal intensity edema (arrowheads). B) Corresponding ﬁ ndings on coronal T2 fat suppressed MR image. Fracture line (arrow), epiphyseal line (small arrow), high signal marrow edema (white arrowheads).

58 Figure 17. 21-year-old conscript suffering from pain in the medial portion of the right ankle for three months. Coronal T2 weighted fat suppressed MR image demonstrates a focal high signal bone marrow edema at the medial malleolus (arrow).

Figure 18. 21-year-old male conscript suffering from right heel pain for 42 days. A) Sagittal T2 weighted fat suppressed MR image depicts a vertical linear area of low signal intensity (arrow) extending through the posterior calcaneus consistent with stress fracture. The adjacent high signal marrow edema (arrowheads) is broad. B) Fracture line (arrow) with surrounding marrow edema (arrowheads), and soft- tissue edema (small arrow) seen on axial T2 weighted fat suppressed MR image.

59 Figure 19. 19-year old male conscript suffering from bilateral ankle pain for 20 days. A) Sagittal T2 weighted fat suppressed MR image demonstrates a fracture line (arrow) with surrounding high signal marrow edema in the head of the talus (in the right leg). Note the anterior soft-tissue edema (small arrow). B) Sagittal T1 weighted MR image shows fracture line in the talar collum (arrow). In talar bone stress injuries, fracture line is parallel to the talonavicular joint (arrowheads).

60 In the tarsal and the metatarsal bones, except for the calcaneus and for the fi fth metatarsal bone, a bone stress injury signifi cantly more often affected several bones. A talus bone stress injury occurred signifi - cantly more often simultaneously with stress injuries to the other bones than as solitary (p=0.007); with navicular bone stress injury (n=22/55, 40%) (p<0.001), and with calcaneal bone stress injury (n=18/55, 33%) (p=0.033). Difference between occurrence of isolated and non-isolated bone stress injuries in the calcaneus failed to reach a statistical signif- icance (p=0.779). A tarsal bone stress injury appeared in combination with stress injuries to other tarsal bones, and with metatarsal bone stress injuries. A metatarsal bone stress injury, in turn, occurred combined with stress injuries of the other metatarsals as well as with tarsal bone stress injuries (Table 7). Pain symptoms related to bone stress injuries of the ankle and/or foot appeared median 30 days (range 3-330 days) from the time of entering the military service. It also became evident that the ankle and/or foot pain in patients with bilateral bone stress injuries appeared sooner (median 14 days, range 2-60 days) than in patients with bone stress injuries in one foot (median 30 days, range 2-330 days) (p=0.027). The duration of the ankle and/or foot pain before a positive fi nding on MRI was median 65 days.

61 V MT 95 IV MT III MT II MT I MT III CUN II CUN OTHER BONES INVOLVED BSI NO OF BSI NO OF 353 47 306 TOTAL 1 >2 TIBIA FIBULA CALC NAVIC CUBOID CUN I Location of bone stress injuries the ankle and foot. IV METATARSALV METATARSAL 17 10 2^# = p<0.001 * = p<0.01 3^ = p<0.05 15^ I-V BONES CUBOID=CUBOIDEUM, CUN I-III=CUNEIFORM I-III BONES, MT I-V=METATARSAL NAVICULAR, CALC=CALCANEUS, NAVIC=TARSAL 7 0 0 0 0 TALUSCALCANEUS NAVICULAR CUBOIDEUM 33CUNEIFORM I 55 33CUNEIFORM II 18 13 13*CUNEIFORM III 36 1 #I METATARSAL 26 1* 20 42*II METATARSAL 17 4 # 33 #III METATARSAL 1 # 29 17* 6 2 42 1* 32 # 3 37 2 # 26 # 0 2 7* 0 1 17* 0 2 # 0 27 # 0 18 35* 0 0 35 # 0 0 22 0 0 8 0 0 8 0 7 1 9 10 3 9 17 5 15 5 8 12 2 6 10 11 17 6 2 11 8 15 2 7 12 8 19 3 14 2 5 8 18 13 1 1 17 3 9 16 0 6 9 15 6 11 4 5 3 14 12 5 15 5 8 4 29 4 7 4 INJURY (BSI) INJURY SITE OF THE BONE STRESS TABLE 7. TABLE

62 6. Discussion

6.1. Patients and study design

The service area of the Central Military Hospital in Helsinki covers 75% of the conscripts in Finland. The conscripts of this study were undergoing military training in the Finnish Defence Forces. The age-range of the patients was narrow. In the prospective study I, recruits came from the elite military conscript unit. The group of conscripts in the retrospective studies II-IV were representative of the general conscript population during their military service. The military equipment in the training in studies II-IV was identical for all conscripts. The patients of the studies II-IV had a history of exercise-induced lower extremity pain disturbing their military service.

6.2. The occurrence and clinical signiﬁ cance of asymptomatic bone stress injuries

In the study I, endosteal marrow edema was counted as a bone stress injury. Bone marrow edema can occur in asymptomatic, physically active people (Lazzarini et al. 1997, Lohman et al. 2001, Kiuru et al. 2003). De- pending on the study design of the previous studies, this marrow edema has been considered either as bone stress injury or as accelerated remodeling rather than a pathological event (Anderson and Greenspan 1996). In the study I, nearly all asymptomatic grade I bone stress injuries in physically active individuals healed, remained as low grade bone stress injuries, or remained asymptomatic even though an intensive physical activity was continued. In our work, grade I bone stress injuries did not progress, and were, consequently, of no clinical signiﬁ cance. In the study

63 I, a relatively high proportion of bone stress injuries (60%) were not associated with symptoms; those of the tibia and the metatarsals tended to be more symptomatic and those of the femoral shaft asymptomatic. In the prospective study of Milgrom et al., 69% of femoral stress fractures but only 8% of tibial stress fractures were asymptomatic (Milgrom et al. 1985). As previously reported, symptoms may depend on the bone involved and the MRI grade of the bone stress injury (Milgrom et al. 1985, Kiuru et al. 2003b). In the study I, the higher grade bone stress injuries tended to be more symptomatic than those of the lower grades. In a literature review, we were unable to ﬁ nd any follow-up study similar to ours.

6.3. Incidence of bone stress injuries

In the present study, the incidence of bone stress injuries was relatively high compared to previous studies. Depending on the location of the bone stress injuries, 1-2/1000 developed as a clinically signiﬁ cant bone stress injury of the lower extremity (study II 1.03/1000, study III 1.99/1000 and IV 1.26/1000). Previous studies concerning knee bone stress injuries have been mainly case reports and, consequently, their real incidence is unknown. The high incidences of bone stress injuries in the knee area (II) and in the femoral bone (III) were probably due to the high sensitivity and speciﬁ city of MRI compared to radiography (Frederickson et al. 1995). When MRI is as widely used as at the institution of the authors to reveal exercise-induced pain related to the musculoskeletal system, also the early stages of bone stress injuries are revealed, which may explain the differences in both the number and the anatomic distribution. As compared with the previous studies, the incidence of fatigue bone stress injury behind exercise-induced ankle and/or foot pain (IV) was relatively low in physically active young adults (Meurman et al. 1980, Fullerton and Snowdy 1988, Korpelainen et al. 2001). It is possible, however, that bone stress injuries of the ankle and foot, especially those of the metatarsals, were already diagnosed by plain radiographs at the primary military health care units, which would explain why they were never sent to orthopaedic consultation and subsequent MRI examination despite prolonged ankle and foot pain (Milgrom and Friedman 2001).

64 6.4. Location and nature of bone stress injuries

In the study II, the medial tibial plateau was also the most common anatomic location involved with bone stress injury (II). The weight-bearing stress in the proximal tibia is thought to be greatest in the medial and posterior parts of the bone (Pentecost et al. 1964). However, in more than half of the cases, the location was exhibited by a fracture line surrounded by bone marrow edema. In a surprisingly high proportion, almost every tenth of conscripts, a bone stress injury in the knee emerged as the cause of exercise-induced knee pain. After commencement of the military training, the patients with bone stress injuries in the medial tibial plateau sought medical advice about two weeks earlier than patients with bone stress injuries in other anatomic locations (II). In a study by Engber (Engber 1977), the delay between the inception of knee pain and the time when medical advice was sought was even shorter, an average of nine days (range 3 to 50). About one third of the patients were discovered to exhibit two bone stress injuries in the same knee simultaneously. This seems to represent a very unusual combination of bone stress injuries in previously healthy individuals, since we found no similar references in the English-language literature. In the study III, the femoral neck was the most common location for bone stress injury, accounting for one-half of all cases. The neck-shaft angle was observed to be normal, 125°-135°, in all patients with a femoral neck bone stress injury. In previous studies, however, a varus alignment of the hip producing a focal concentration of mechanical stress in the femoral neck, and alterations in the normal hip biomechanics have appeared to be associated with femoral neck bone stress injuries (Carpin- tero et al. 2003, Hohmann et al. 2004). Other anatomic variations, such as unequal leg length, have been suggested to occasionally predispose to fatigue fractures (Blickenstaff and Morris 1966, Friberg 1982). In the study III, the proximal femoral bone stress injuries were as common as previously reported in athletes (Hershman et al. 1990a, Johnson et al. 1994, Anderson et al. 2001), those of the middle-distal shaft were rare. It is possible that proximal shaft bone stress injuries existed in the past without being diagnosed, because they were less likely to become displaced (Provost and Morris 1969, Johnson et al. 1994). For the same reason, the proportion of displaced distal shaft stress fractures in older

65 studies seems high (Provost and Morris 1969, Hallel et al. 1976). In the present study, the duration of pain was shorter in patients with femoral neck bone stress injury or proximal femoral shaft bone stress injury than in patients with femoral condylar bone stress injury, and the higher the bone stress injury grade, the shorter the duration of pain. The location of pain was typically verifi ed to be in the site of a bone stress injury in this study. In the study III, the relative proportion of female patients with femoral neck bone stress injury was greater than that of males. This may be explained by female-specifi c intrinsic and extrinsic factors, but also by heavy military equipment that is similar for both sexes. Previous studies in athletes have not shown differences between males and females (Luchini et al. 1983, Lloyd-Smith et al. 1985). In the study IV, the number of cases with a tarsal bone stress injury was relatively high; more than a half of all bone stress injuries occurred in the tarsal bones. More than one-third of all bone stress injuries occurred in the metatarsal bones. The second and third metatarsal bones were the most commonly affected, a fi nding similarly described in previous studies (Orava et al. 1978, Meurman 1981a, Resnick and Kang 1997, Boden et al. 2001, Mäenpää et al. 2002). More than half of the bone stress injuries of the ankle and foot occurred as multiple injuries, which is in concordance with previous studies (Meurman et al. 1980, Korpelainen et al. 2001). The talus showed signifi cantly more often combined bone stress injuries than isolated ones. In contrast, the combination of a calcaneal bone stress injury with injuries to adjacent bones was not statis- tically signifi cant. According to previous studies based on radiographs and scintigram, bone stress injuries of the talus, the cuboideum, and the cuneiform bones occurring both in military recruits and athletes are rare whereas calcaneus is the tarsal bone with the highest number of stress fractures (Orava 1980, Meurman 1981a, Hershman and Mailly 1990b, Resnick and Kang 1997, Weber et al. 2005). Calcaneal bone stress injuries have been encountered in activities like parachute jumping and in prolonged standing among military trainees and athletes (Hershman and Mailly 1990b).

66 6.5. Imaging of bone stress injuries

In the present study, the results of radiography fi ndings were not compared with MR fi ndings but the main focus was on MRI. Radiographs are indicated for the fi rst imaging examination whenever there is a suspicion of a stress fracture or overuse injury (Fredericson et al. 1995, Ander- son and Greenspan 1996, Bergman and Fredericson 1999). As mentioned earlier, magnetic resonance imaging is an important imaging modality for the musculoskeletal system, since various tissues display distinctive signal intensities on T1- and T2–weighted images. Fat suppression techniques, T2-weighted images with fat suppression and STIR images, were the most commonly used sequences for detecting increased amount of extracellular water seen in bone stress and in more advanced bone stress injuries. MRI is also the method of choise in revealing trabecular bone changes. (Stafford 1986, Schweitzer 1986, Vogler and Murphy 1988, Lee and Yao 1988, Kiuru 2002). Since adolescents have a relatively intense metabolic activity in the epiphyseal and metaphyseal areas (Geslien et al. 1976), a comparative study requires MR imaging of both legs. However, in case of a “mild” marrow edema (Gr I) around the epiphyseal line of the knee, the lack of bilateral MR images in the present study made comparing impossible. Epiphyseal lines should be closed at the approximate age of 20-22 in males and a few years earlier in females. Theoretically, after the closure of the epiphyses, these areas may remain weaker and therefore vulnerable to bone stress injuries. That only painful sides were imaged or included in studies (II-IV), it is unsure whether bilateral bone stress injuries occurred or not. Moreover, the knowledge that bone tumors, albeit rare, favor the knee area in this age group of young patients, calls for accurate diagnosis between bone stress injuries and other musculoskeletal entities.

67 7. Conclusions

On the basis of the present results, the following conclusions can be drawn:

Study I A total of 75 bone stress injuries were found in 16 out of a total of 21 elite-unit military recruits. Asymptomatic grade I bone stress injuries were common in subjects undergoing intensive physical training. Only 40% of the bone stress injuries were symptomatic. Symptoms depended on the location and MRI grade of injury, with higher grades usually more symptomatic. Repeated clinical and MRI assessment indicated that 84% of asymptomatic grade I bone stress injuries healed or remained as grade I and asymptomatic even when the intensive physical activity continued.

Study II The incidence of bone stress injuries in the knee underlying exercise- induced knee pain seems to be relatively high, 103 per 100,000 person- years (88 patients out of a total of 85,318 conscript-years, a total of 1330 patients MR imaged, 141 bone stress injuries) in Finnish conscripts. Two solitary bone stress injuries in the same knee simultaneously occurred in 28% of the patients. The most common anatomic location was the medial tibial plateau, which was also the most typical location for a more advanced injury. After commencement of the military service, a bone stress injury in the medial tibial plateau caused knee pain earlier than a bone stress injury elsewhere in the knee (p=0.014). When addressing a patient-history typically indicating no recent trauma, unclear ﬁ ndings in the physical examination, and negative plain radiographs, MRI should be emphasized as the diagnostic method for bone stress injuries in the knee before arthroscopy is performed.

68 Study III Fatigue femoral bone stress injuries seem to be relatively common in physically active patients. The incidence of femoral bone stress injuries was 199 per 100,000 person-years (170 patients out of a total of 85,318 conscripts-years, a total of 1857 patients MR imaged, 185 bone stress injuries). The three most common sites were the neck (50%), the condylar area (24%), and the proximal shaft (18%). A stress reaction occurred in 57% and a fracture line in 22% of the cases. There was a statistical correlation between the bone stress injury grade and the duration of pain (p=0.001). Bone stress injuries in the upper femur showed a shorter duration of pain than bone stress injuries in the condylar area of the knee. The location of pain was usually situated in the site of the bone stress injury.

Study IV Multiple various-stage bone stress injuries of the ankle and foot were found in physically active young adults experiencing exercise-induced ankle and/or foot pain during military service. The incidence of bone stress injuries in the ankle and foot was 126 per 100,000 person-years (131 patients with a positive MR fi nding out of 104,340 conscript-years, a total of 267 patients MR imaged, 378 bone stress injuries). The most commonly affected anatomic areas were the tarsal bones (58%) followed by the metatarsal bones (36%). Multiple bone stress injuries in one foot occurred in 63% of the cases. In tarsal bones except the calcaneus, and in metatarsal bones except the fi fth metatarsal bone, a bone stress injury was signifi cantly more often seen in more than one bone.

69 8. Summary

The conscripts of this study were undergoing military training in the Finnish Defence Forces. In study I, recruits came from the elite military conscript unit, the Parachute School of the Utti Jaeger Regiment. The group of conscripts included in the studies II-IV is representative of the general conscript population. Military equipment used in the training was identical for all patients in studies II-IV. All patients in studies II-IV had a history of exercise-induced lower extremity pain disturbing their military service. The purpose of the study I was to investigate the occurrence of asymptomatic bone stress injuries by clinical and MRI follow-up, their clinical signiﬁ cance, and whether they progress to stress fractures in recruits undergoing intensive physical training. The elite-unit military recruits (I) were examined by clinical and MRI follow-up before commencement of the intensive training period, at 6 weeks from commencement, and on completion of the 5-month training program. In three original articles (II-IV) based on MRI, the incidence, location, nature, and patterns of bone stress injuries in the knee, the femoral bone, and the ankle and foot in physically active military conscripts during their military service were described. In all studies (I-IV), the MR images with bone stress injury ﬁ ndings were evaluated with respect to bone stress injury location and bone stress injury type, and bone stress injuries were graded from endosteal marrow edema (Grade I) to callus in cortical bone (Grade V). Marrow edema represented as low or intermediate signal intensity on T1-weighted images and as high signal intensity on T2-weighted and STIR images. A fracture line appeared as a low signal intensity line on all of these pulse sequences. In the study I, based on three MRI scans, a total of 75 bone stress injuries were found in 16 out of a total of 21 recruits, of which only 40% were symptomatic. Symptoms depended on the location and MRI grade of injury, with higher grades usually more symptomatic. Repeated clinical and MRI assessment indicated that 84% of asymptomatic grade I

70 bone stress injuries healed or remained as grade I and asymptomatic. In the study II, the incidence of bone stress injuries in the knee was 103 per 100,000 person-years (88 patients with a positive MR fi nding out of 85,318 conscript-years, 141 bone stress injuries, a total of 1330 patients MR imaged). Bilateral bone stress injuries in the knee were found in 25% and two solitary bone stress injuries in the same knee simultaneously in 28% of the patients. The medial tibial plateau was the most common anatomic location and the most typical location for a more advanced injury. It also caused knee pain earlier than a bone stress injury elsewhere in the knee (p=0.014) after the commencement of the military service. In the study III, the incidence of femoral bone stress injuries was 199 per 100,000 person-years (170 patients with a positive MR fi nding out of 85,318 conscript-years, 185 bone stress injuries, a total of 1857 patients MR imaged). Bilateral injuries occurred in 9% of the cases. The three most common sites were the neck (50%), the condylar area (24%), and the proximal shaft (18%). A stress reaction occurred in 57% and a fracture line in 22% of the cases. There was a statistical correlation between the bone stress injury grade and the duration of pain (p=0.001). In the study IV, the incidence of bone stress injuries in the ankle and foot was 126 per 100,000 person-years (131 patients with a positive MR fi nding out of 104,340 conscript-years, 378 bone stress injuries, a total of 267 patients MR imaged). Of the injuries, 57.7% occurred in the tarsal bones of which the talus and the calcaneus were the most prominent single bones and the most common locations for higher grade (Gr IV- V) bone stress injuries. 35.7% of the bone stress injuries occurred in the metatarsal bones. Multiple bone stress injuries in one foot occurred in 63% of the cases. In tarsal bones except the calcaneus, and in metatarsal bones except the fi fth metatarsal bone, a bone stress injury was signifi - cantly more often seen in more than one bone. Most of the cases were low grade (88%) bone stress injuries (Gr I- III). In conclusion, asymptomatic grade I bone stress injuries are common in subjects undergoing intensive physical training and most of them heal or remain as asymptomatic grade I bone stress injuries even when the intensive physical activity continues. The incidence of bone stress injuries in the knee underlying exercise-induced knee pain seems to be relatively high in conscripts. In the femur, the most common anatomic site was the neck followed by the condylar area and the proximal shaft. Fatigue femoral bone stress injuries seem to be relatively common in physical-

71 ly active patients. The location of pain is usually situated in the site of the bone stress injury. Bone stress injuries in the upper femur showed a shorter duration of pain than bone stress injuries in the condylar area of the knee. Multiple various-stage bone stress injuries of the ankle and foot were found in physically active young adults behind exercise-induced ankle and/or foot pain during military service. Most commonly affected was the anatomic area occupying the tarsal bones.

72 9. Acknowledgements

The present study was carried out at the Research Institute of Military Medicine, Central Military Hospital. I owe my deepest gratitude to my supervisors, Docent Harri Pihla- jamäki, and Docent Martti Kiuru, for fi rst introducing this study to me and then guiding me through this research project. I express my sincere gratitude for their valuable advice during the entire work process. I wish to express my gratitude to Professor Leena Kivisaari, Depart- ment of Diagnostic Radiology, University of Helsinki about encouraging me to start and continue the study project. I also express my gratitude to Surgeon General of the Finnish Defence Forces, Brigad Gen Pentti Kuronen (M.C.) and Former Surgeon General, MajGen Timo Sahi (M.C.) for the creative spirit and encouraging attitude to the research work in the fi eld of military medicine. To my offi cial reviewers, Docent Riitta Parkkola and Docent Jari Parkkari, I wish to express my profound gratitude for their constructive criticism and generous help in completing this dissertation. I am also grateful to all my co-authors, Docents Riina Haataja and Juhani Ahovuo, and Drs. Anssi Reponen and Markus Sormaala. I wish to present my sincere gratitude to Dr. Ville Mattila, the Research Institute of Military Medicine, for valuable comments during this study. I express my warmest thanks for Docent Markus Henriksson, the Head of the Central Military Hospital for the help received from the hospital organization for this research. I am grateful to Docent Matti Mäntysaari, the Head of the Research Institute of Military Medicine for the opportu- nity to use the research facilities of the Institute. I wish to express my thanks for all doctors of the Central Military Hospital, especially to Docents Heikki Korpela and Tuomo Visuri, and Drs. Marja Eskelin and Jarmo Skyttä, for their encouraging support and for creating an atmosphere that enabled me to carry out this project. I also wish to thank all the helpful staff of the Central Military Hospital “Tilkka” for arranging me an interesting program during this research study.

73 I want to thank Mr. Harry Larni, Research Scientist, for illustrating this thesis, and Mrs Taina Saarni, Mrs Kirsi Hannula, and Mr Kari Kelho - for their secretarial help and assistance for the thesis. I am very grateful to Mrs Marja Vajaranta for editing the English language in the articles and in the dissertation. This project was made possible by the fi nancial support of The Scien- tifi c Advisory Board for Defence, which I gratefully acknowledge. I warmly thank my parents, Pirkko and Jouko, and my brothers Markus and Ilkka, for their love and support all my life. Without the support of dear friends, Pia Mustakangas et al. I could not have fi nished this thesis. For Emil and Karim, thank you both for the unforgettable moments during this project.

74 10. References

Aleman O: Om s.k. marschvulst (syndesmitis metatarsea). Tidskr Milit Hälsov 54: 191-208, 1929 as quoted in Morris and Blickenstaff 1967.

Almeida SA, Williams KM, Shaffer RA, Brodine SK. Epidemiological patterns of musculoskeletal injuries and physical training. Med Sci Sports Exerc 31(8):1176-82, 1999.

Anderson MW, Greenspan A: Stress fractures. Radiology 199-1:1-12, 1996.

Anderson MW, Kaplan PA, Dussault RG: Adductor insertion avulsion syndrome (thigh splints): spectrum of MR imaging features. AJR Am J Roentgenol 177-3:673-5, 2001.

Arendt EA, Grifﬁ ths HJ: The use of MR imaging in the assessment and clinical management of stress reactions of bone in high-performance athletes. Clin Sports Med 16-2:291-306, 1997.

Asal: Überlastungsschäden am Knochensystem bei Soldaten. Arch Klin Chir 186: 511-522, 1936 as quoted in Morris and Blickenstaff 1967.

Asal: Über Entstehung und Verhütung der Spontanfrakturen an den unteren Extremitä- ten. Veröffentl. Gebiete Heeres-Sanitätsw. 104:132-175, 1937 as quoted in Morris and Blickenstaff 1967.

Beck TJ, Ruff CB, Mourtada FA, Shaffer RA, Maxwell-Williams K, Kao GL, Sartoris DJ, Brod- ine S: Dual-energy X-ray absorptiometry derived structural geometry for stress fracture prediction in male U.S. Marine Corps recruits. J Bone Miner Res 11(5):645-53, 1996.

Bennell KL, Malcolm SA, Thomas SA, Reid SJ, Brukner PD, Ebeling PR, Wark JD: Risk factors for stress fractures in track and ﬁ eld athletes. A twelve-month prospective study. Am J Sports Medc24(6):810-8, 1996a.

Bennell KL, Malcolm SA, Thomas SA, Wark JD, Brukner PD: The incidence and distribution of stress fractures in competitive track and ﬁ eld athletes. A twelve-month prospective study. Am J Sports Med 24: 211-217, 1996b.

Bennell KL, Brukner PD: Epidemiology and site speciﬁ city of stress fractures. Clin Sports Med 16-2:179-96, 1997.

Bergman AG, Fredericson M: MR imaging of stress reactions, muscle injuries, and other overuse injuries in runners. Magn Reson Imaging Clin N Am 7: 151-174, 1999.

Blecher A: Über den Einﬂ uss des Parademarsches auf die Entstehung der Fussgeschwulst. Med Klin 13:305, 1905 as quoted in Morris and Blickenstaff 1967.

Blickenstaff LD, Morris JM: Fatigue fracture of the femoral neck. J Bone Joint Surg 48A: 1031-1047, 1966 as quoted by Fullerton LR Jr and Snowdy HA in Femoral neck stress fractures. Am J Sports Med 16-4:365-77, 1988.

75 Boden BP, Speer KP: Femoral stress fractures. Clin Sports Med 16: 307-317, 1997.

Boden BP, Osbahr DC: High-risk stress fractures: evaluation and treatment. J Am Acad Or- thop Surg 8-6:344-53, 2000.

Boden BP, Osbahr DC, Jimenez C: Low-risk stress fractures. Am J Sports Med 29: 100-111, 2001.

Breithaupt J: Zur pathologie des menschlichen fussess. Medizin Zeitung 24:169-177, 1855 as quoted by Matheson et al. 1987.

Buckley SL, Robertson WW Jr, Shalaby-Rana E: Stress fractures of the femoral diaphysis in young children. A report of 2 cases. Clin Orthop Relat Res (310):165-169, 1995.

Buckwalter JA, Glimcher MJ, Cooper RR, Recker R: Bone biology II: Formation, form, modeling, remodeling, and regulation of cell function. Instr Course Lect 45:387-399, 1996.

Burr DB, Milgrom C, Boyd RD, Higgins WL, Robin G, Radin EL: Experimental stress fractures of the tibia. Biological and mechanical aetiology in rabbits. J Bone Joint Surg Br 72-3:370-5, 1990.

Burrows HJ: Fatigue fractures of the ﬁ bula. J Bone Joint Surg 30-B:266-279, 1948.

Carpintero P, Leon F, Zafra M, Serrano-Trenas JA, Roman M: Stress fractures of the femoral neck and coxa vara. Arch Orthop Trauma Surg 123-6:273-7, 2003.

Carter DR, Caher WE, Spengler DM, Frankel VH: Uniaxial fatigue of human cortical bone. The inﬂ uence of tissue physical characteristics. B Biomech 14:461-470, 1981.

Clement DB, Ammann W, Taunton JE, Lloyd-Smith R, Jesperson D, McKay H, Goldring J, Matheson GO: Exercise-induced stress injuries to the femur. Int J Sports Med 14-6:347- 52, 1993.

Daffner RH: Stress fractures: Current concepts. Skeletal Radiol 2: 221-229, 1978.

Daffner RH: Anterior tibial striations. AJR 143:651-653, 1984.

Daffner RH, Pavlov H: Stress fractures: current concepts. Am J Roentgenol 159: 245-252, 1992.

Damadian R: Tumor detection by nuclear magnetic resonance. Science 171: 1151-1153, 1971.

Damadian R, Goldsmith M, Minkoff L: NMR in Cancer: Fonar image of the live human body. Physiol Chem Phys 8:97-108, 1977.

Devas MB, Sweetnam R: Stress fractures of the ﬁ bula: A review of ﬁ fty cases in athletes. J Bone Joint Surg 38-B:818-829, 1956.

Devas MB: Stress fractures of the femoral neck. J Bone Joint Surg Br 47-4:728-38, 1965.

Devas M : Stress fractures. Edinburgh, Scotland : Churchill Livingstone, 1975 as quoted by Anderson MW and Greenspan A 1996.

Egol KA, Koval KJ, Kummer F, Frankel VH: Stress fractures of the femoral neck. Clin Orthop Relat Res 348:72-8, 1998.

Engber WB: Stress fractures of the medial tibial plateau. J Bone Joint Surg 59A: 767-769,

76 1977.

Erne P, Burckhardt A: Femoral neck fatigue fracture. Arch Orthop Trauma Surg 97-3:

213-220, 1980.

Fazzalari NL: Trabecular microfracture. Calcif Tissue Int 53 Suppl 1:S143-6, 1993.

Fredericson M, Bergman AG, Hoffman KL, Dillingham MS: Tibial stress reaction in runners. Correlation of clinical symptoms and scintigraphy with a new magnetic resonance imaging grading system. Am J Sports Med 23: 472-481, 1995.

Friberg O: Leg length asymmetry in stress fractures. A clinical and radiological study. J Sports Med 22: 485, 1982.

Frost HM: Some ABC’s of skeletal pathophysiology III: bone balance and the ”B.BMU. Edi- torials. Calcif Tissue Int 45: 131-133, 1989a.

Frost HM: The biology of fracture healing. An overview for clinicians. Part I-II. Clin Orthop Relat Res 248: 283-293, 1989b.

Frost HM: Structural adaptations to mechanical usage (SATMU): 1. Redeﬁ ning Wolff’s law: The bone modeling problem. Anat Rec 226:403-413, 1990a.

Frost HM: Structural adaptations to mechanical usage (SATMU): 2. Redeﬁ ning Wolff’s law: The bone modeling problem. Anat Rec 226:414-422, 1990b.

Frost HM: Some ABC’s of skeletal pathophysiology. 6. The growth / modeling / remodeling distinction. Calcif Tissue Int 49: 301-302, 1991.

Fullerton LR Jr, Snowdy HA: Femoral neck stress fractures. Am J Sports Med 16-4:365-77, 1988.

Genz: Vermehrtes Auftreten der alten Marsch- und Fussgeschwulst. Veröffentl. Gebiete Heeres-Sanitätsw. 98:58, 1935 as quoted in Morris and Blickenstaff 1967.

Geslien GE, Thrall JE, Espinosa JL, Older RA: Early detection of stress fractures using Tc-poly- phosphate. Radiology 121: 683-687, 1976.

Giladi M, Ahronson Z, Stein M, Danon YL, Milgrom C: Unusual distribution and onset of stress fractures in soldiers. Clin Orthop Relat Res 192:142-6, 1985.

Giladi M, Milgrom C, Simkin A, Danon Y: Stress fractures. Identiﬁ able risk factors. Am J Sports Med. 19(6):647-52, 1991.

Gilbert RS, Johnson HA: Stress fractures in military conscripts – a review of twelve years’ experience. Mil Med 131:716-721, 1966.

Glorioso JE, Ross G, Leadbetter WB, Boden BP: Femoral supracondylar stress fractures. An unusual cause of knee pain. Physician Sportsmed 30:25-9, 2002.

Greaney RB, Gerber FH, Laughlin RL, Kmet JP, Metz CD, Kilcheski TS, Rao BR, Silverman ED: Distribution and natural history of stress fractures in U.S. Marine recruits. Radiology 146: 339-346, 1983.

Hallel T, Amit S, Segal D: Fatigue fractures of tibial and femoral shaft in soldiers. Clin Or- thop Relat Res 118:35-43, 1976.

77 Hensal F, Nelson T, Pavlov H, Torg JS: Bilateral patellar fractures from indirect trauma. A case report. Clin Orthop 178: 207-209, 1983.

Hershman EB, Lombardo J, Bergfeld JA: Femoral shaft stress fractures in athletes. Clin Sports Med 9-1:111-9, 1990a.

Hershman EB, Mailly T: Stress fractures. Clin Sports Med 9(1):183-214, 1990b.

Hohmann E, Wortler K, Imhoff AB: MR imaging of the hip and knee before and after marathon running. Am J Sports Med 32: 55-59, 2004.

Hulkko A, Orava S: Stress fractures in athletes. Int J Sports Med 8(3):221-6, 1987.

Hyman AA, Heiser WJ, Kim SE, Norfray JF: An excavation of the distal femoral metaphysis: a magnetic resonance imaging study. A case report. J Bone Joint Surg (Am) 77: 1897- 901, 1995.

Johnson AW, Weiss CB Jr, Wheeler DL: Stress fractures of the femoral shaft in athletes-- more common than expected. A new clinical test. Am J Sports Med 22-2:248-56, 1994.

Johnson LC, Stradford HT, Geis RW, Dineen JR, Kerley E: Histogenesis of stress fractures. J Bone Joint Surg Am 45:1542, 1963.

Jones BH, Harris JM, Vinh TN, Rubin C: Exercise-induced stress fractures and stress reactions of bone: epidemiology, etiology, and classiﬁ cation. Exerc Sport Sci Rev 17:379- 422, 1989.

Jones BH, Cowan DN, Tomlinson JP, Robinson JR, Polly DW, Frykman PN: Epidemiology of injuries associated with physical training among young men in the army. Med Sci Sports Exerc 25(2):197-203, 1993.

Kanstrup IL: Bone scintigraphy in sports medicine: a review. Scand J Med Sci Sports 7(6):322- 330, 1997.

Kelly EW, Jonson SR, Cohen ME, Shaffer R: Stress fractures of the pelvis in female navy recruits: an analysis of possible mechanisms of injury. Mil Med 165(2):142-6, 2000.

Kiuru MJ, Pihlajamäki HK, Perkiö JP, Ahovuo JA: Dynamic contrast-enhanced MR imaging in symptomatic bone stress of the pelvis and the lower extremity. Acta Radiol 42-3:277- 85, 2001.

Kiuru MJ, Pihlajamäki HK, Hietanen HJ, Ahovuo JA: MR imaging, bone scintigraphy, and radiography in bone stress injuries of the pelvis and the lower extremity. Acta Radiol 43: 207-212, 2002.

Kiuru MJ, Mäntysaari MJ, Pihlajamäki HK, Ahovuo JA: Evaluation of stress-related anterior lower leg pain with magnetic resonance imaging and intracompartmental pressure measurement. Mil Med 168(1): 48-52, 2003a.

Kiuru MJ, Pihlajamäki HK, Ahovuo JA: Fatigue stress injuries of the pelvic bones and proximal femur: evaluation with MR imaging. Eur Radiol 13(3):605-11, 2003b.

Korpelainen R, Orava S, Karpakka J, Siira P, Hulkko A: Risk factors for recurrent stress fractures in athletes. Am J Sports Med 29(3): 304-10, 2001.

Krivickas LS: Anatomical factors associated with overuse sports injuries. Sports Med 24-2:132-46, 1997.

78 Kuusela TV: Stress fracture. A radionuclide, roentgenological and clinical study of Finnish conscripts. Annales Medicinae Militaris Fenniae, 1980.

Lassus J, Tulikoura I, Konttinen YT, Salo J, Santavirta S: Bone stress injuries of the lower extremity: a review. Acta Orthop Scand 73-3:359-68, 2002.

Lazzarini KM, Troiano RN, Smith RC: Can running cause the appearance of marrow edema on MRI of the foot and ankle? Radiology 202:540-542, 1997.

Lee JK, Yao L: Stress fractures: MR imaging. Radiology 169: 217-220, 1988.

Li GP, Zhang S, Chen C, Chen H, Wang AM: Radiographic and histologic analyses of stress fracture in rabbit tibias. Am J Sports Med 13: 285-295, 1998.

Lohman M, Kivisaari A, Vehmas T, Kallio P, Malmivaara A, Kivisaari L: MRI abnormalities of foot and ankle in asymptomatic, physically active individuals. Skeletal Radiol 30-2:61-6, 2001.

Loyd-Smith R, Clement DB, McKenzie DC, Taunton JE: A Survey of overuse and traumatic hip and pelvic injuries in athletes. Physician Sportsmed 13:8-10, 1985.

Luchini MA, Sarokhan AJ, Micheli LJ: Acute displaced femoral-shaft fractures in long-distance runners. Two case reports. J Bone Joint Surg Am 65-5:689-91, 1983.

Manaster BJ, Disler DG, May DA (eds.): Musculoskeletal Imaging: The Requisites. Mosby Inc., Missouri, 2002.

Mason RW, Moore TE, Walker CW, Kathol MH: Patellar fatigue fractures. Skeletal Radiol- ogy 25: 329-332, 1996.

Matheson GO, Clement DB, McKenzie DC, Taunton JE, Lloyd-Smith DR, MacIntyre JG: Stress fractures in athletes: A study of 320 cases. Am J Sports Med 15: 46-58, 1987.

McBryde AM Jr: Stress fractures in athletes. J Sports Med 3(5):212-7, 1975.

McBryde AM Jr: Stress fractures in runners. Clin Sports Med 4-4:737-52, 1985.

Meurman KO, Elfving S: Stress fractures in soldiers: A multifocal disorder. Radiology 134: 483-487, 1980a.

Meurman KO, Elfving S: Stress fracture of the cuneiform bones. Br J Radiol 53: 157-160, 1980b.

Meurman KO: Less common stress fractures in the foot. Br J Radiol 54:1-7, 1981a.

Meurman KO: Stress fractures in soldiers. An analysis of 986 consecutive cases. Acta Univ. Oul. D 78 Radiol 3, 1981b.

Meurman KO, Somer K, Lamminen A: Stressfracturen des femurs bei soldaten. Fortschr Röntgenstr 134-5:528-32, 1981c.

Meyers SP, Wiener SN: Magnetic resonance imaging features of fractures using the short tau inversion recovery (STIR) sequence: correlation with radiographic ﬁ ndings. Skeletal Radiol 20: 499-507, 1991.

Milgrom C, Giladi M, Stein M, Kashtan H, Margulies JY, Chisin R, Steinberg R, Aharonson Z: Stress fractures in military recruits: a prospective study showing an unusually high incidence. J Bone Joint Surg 67B: 732-735, 1985.

79 Milgrom C, Chisin R, Margulies J, Giladi M, Stein M, Kashtan H, Atlan H: Stress fractures of the medial femoral condyle. J Trauma 26: 199-200, 1986.

Milgrom C, Finestone A, Shlamkovitch N, Rand N, Lev B, Simkin A, Wiener M: Youth is a risk factor for stress fracture. A study of 783 infantry recruits. J Bone Joint Surg 76B: 20-22, 1994.

Milgrom CE, Friedman E: Early diagnosis and clinical treatment of stress fractures. In: Burr DB, Milgrom C (eds.): Musculoskeletal Fatigue and Stress Fractures. CRC Press, 2001.

Morris JM, Blickenstaff LD: Fatigue Fractures: A Clinical Study. Charles C Thomas, Spring- ﬁ eld IL 1967.

Muller W: Der militärische Abrissermudungsschaden. Dtsch militärarzt 8: 283-286, 1943

Mulligan ME: The “gray cortex”: an early sign of stress fracture. Skeletal Radiol 24: 201- 203, 1995.

Muralikuttan KP, Sankarart-Kutty M: Supracondylar stress fracture of the femur. Injury 30: 66-67, 1999.

Mäenpää H, Lehto MU, Belt EA: Stress fractures of the ankle and forefoot in patients with inﬂ ammatory arthritides. Foot Ankle Int 23(9): 833-837, 2002.

Nielsen MB, Hansen K, Holmer P, Dyrbye M: Tibial periosteal reactions in soldiers. A scintigraphic study of 29 cases of lower leg pain. Acta Orthop Scand 62: 531-534, 1991.

Ollonqvist LJ: Kallusmuodostumista sääriluussa ilman luunmurtumaa. Duodecim 45:474- 486, 1929.

Ollonqvist LJ: Osteopathia itineraria tibiae. Acta Radiol 18: 526, 1937.

Orava S, Puranen J, Ala-Ketola L: Stress fractures caused by physical exercise. Acta Orthop Scand 49(1):19-27, 1978.

Orava S: Stress fractures. Br J Sports Med 14-1:40-4, 1980.

Orava S, Taimela S, Kvist M, Karpakka J, Hulkko A, Kujala U: Diagnosis and treatment of stress fracture of the patella in athletes. Knee Surg Sports Traumatol Arthrosc 4: 206- 211, 1996.

Ozburn MS, Nichols JW: Pubic ramus and adductor insertion stress fractures in female basic trainees. Mil Med 146-5:332-4, 1981.

Pathria MN, Isaacs P: Magnetic resonance imaging of bone marrow. Curr Opin Radiol 4: 21-31, 1992.

Peltokallio P: Tyypilliset urheiluvammat I ja II. Vammalan kirjapaino, Vammala 2003.

Pentecost RL, Murray RA, Brindley HH: Fatigue, insufﬁ ciency and pathologic fractures. JAMA 187: 1001-1004, 1964.

Prather JL, Nusynowitz ML, Snowdy HA, Hughes AD, McCartney WH, Bagg RJ: Scintigraphic ﬁ ndings in stress fractures. J Bone Joint Surg Am 59:869-874, 1977.

Provost RA, Morris JM: Fatigue fracture of the femoral shaft. J Bone Joint Surg 51-A: 487- 498, 1969.

80 Resnick D, Greenway G: Distal femoral cortical defects, irregularities, and excavations. Ra- diology 143(2): 289-295, 1982.

Resnick D, Kang H: Traumatic disorders of bone. In: Internal Derangements of joints. Em- phasis on MR Imaging. W.B. Saunders Company, Pihladelphia, 1997.

Rupani HD, Holder LE, Espinola DA, Engin SI: Three-phase radionuclide bone imaging in sports medicine. Radiology 156(1):187-96, 1985.

Sahi T: Stress fracture: Epidemiology and control. Rev Int Sante Armees 57: 311-313, 1984.

Sahi T, Peitso A, Laapio H: The incidence of stress fractures [In Finnish] In: Koskenvuo K, Sahi T, Peitso A (eds.): Military health and medical care. Pp. 208-211, Karisto Oy, Hämeenlinna, 1996.

Sallis RE, Jones K: Stress fractures in athletes. How to spot this underdiagnosed injury. Post- grad Med 89(6):185-8, 191-2, 1991.

Salminen ST, Pihlajamäki HK, Visuri TI, Böstman OM: Displaced fatigue fractures of the femoral shaft. Clin Orthop Relat Res 409:250-9, 2003.

Sanders TG, Williams PM, Vawter KW: Stress fracture of the tarsal navicular. Mil Med 169: 8-13, 2004.

Savoca CJ: Stress Fractures: a classiﬁ cation of the earliest radiographic signs. Radiology 100: 519-524, 1971.

Schmidt-Brudvig TJ, Gudger TD, Obermeyer L: Stress fractures in 295 trainees: a one-year study of incidence as related to age, sex, and race. Mil Med 148-8:666-7, 1983.

Schweitzer ME, White LM: Does altered biomechanics cause marrow edema? Radiology 198-3:851-3, 1996.

Shaffer RA, Brodine SK, Almeida SA, Williams KM, Ronaghy S: Use of simple measures of physical activity to predict stress fractures in young men undergoing a rigorous physical training program. Am J Epidemiol 149(3):236-42, 1999.

Shin AY, Morin WD, Gorman JD, Jones SB, Lapinsky AS: The superiority of magnetic resonance imaging in differentiating the cause of hip pain in endurance athletes. Am J Sports Med 24(2):168-76, 1996.

Shin AY, Gillingham BL: Fatigue Fractures of the Femoral Neck in Athletes. J Am Acad Or- thop Surg 5-6:293-302, 1997.

Sokoloff MR, Farooki S, Resnick D: Spontaneous osteonecrosis of the knee associated with ipsilateral tibial plateau stress fracture: report of two patients and review of the literature. Skeletal Radiol 30: 53-56, 2001.

Stafford SA, Rosenthal DI, Gebhardt MC, Brady TJ, Scott JA: MRI in stress fracture. AJR Am J Roentgenol 147: 553-556, 1986.

Stechow: Fussödem und Röntgenstrahlen. Deutsche Militärärztliche Zeitschrift 26: 465- 471, 1897.

Steinbach LS: Painful syndromes around the ankle and foot: magnetic resonance imaging evaluation. Top Magn Reson Imaging 5:311-326, 1998.

Visuri T, Vara A, Meurman KO: Displaced stress fractures of the femoral neck in young male

81 adults: a report of twelve operative cases. J Trauma 28-11:1562-9, 1988.

Visuri T, Hietaniemi K: Displaced stress fracture of the femoral shaft: a report of three cases. Mil Med 157-6:325-7, 1992.

Visuri T: Stress osteopathy of the femoral head. 10 military recruits followed for 5-11 years. Acta Orthop Scand 68-2:138-41, 1997.

Vogler JB III, Murphy WA: Bone marrow imaging. Radiology 168: 679-693, 1988.

Volpin G, Petronius G, Hoerer D, Stein H: Lower limb pain and disability following strenuous activity. Mil Med 154-6:294-7, 1989.

Volpin G, Hoerer D, Groisman G, Zaltzman S, Stein H: Stress fractures of the femoral neck following strenuous activity. J Orthop Trauma 4-4:394-8, 1990.

Vossinakis IC, Tasker TPB: Stress fracture of the medial tibial condyle. The Knee 7: 187-190, 2000.

Välimäki VV, Alfthan H, Lehmuskallio E, Löyttyniemi E, Sahi T, Suominen H, Välimäki MJ: Risk factors for clinical stress fractures in male military recruits: A prospective cohort study. Bone, 2005.

Weber JM, Vidt LG, Gehl RS, Montgomery T: Calcaneal stress fractures. Clin Podiatr Med Surg 22(1):45-54, 2005.

Weber PC: Salter-Harris type II stress fracture in a young athlete: A case report. Orthopedics 11: 309-311, 1988.

Wolff J: Das Gesetz der Transformation der Knochen. Verlag von August Hirschwald, Berlin 1892.

Yasuda T, Miyazaki K, Tada K, Matsuda Y, Murakami H: Stress fracture of the right distal femur following bilateral fractures of the proximal ﬁ bulas: A case report. Am J Sports Med 20: 771-774, 1992.

Zlatkin MB, Bjorkengren A, Sartoris DJ, Resnick D: Stress fractures of the distal tibia and calcaneus subsequent to acute fractures of the tibia and ﬁ bula. AJR Am J Roentgenol 149(2):329-32, 1987.

Zwas ST, Elkanovitch R, Frank G: Interpretation and classiﬁ cation of bone scintigraphic ﬁ ndings in stress fractures. J Nucl Med 28(4): 452-457, 1987.