Journal of the American Academy of Orthopaedic Surgeons 344 Stress fractures are common inju- ries frequently seen in athletes and military recruits. Although the re- ported incidence of stress fractures in the general athletic population is less than 1%, the incidence in run- ners may be as high as 20%. In a review of 370 athletes with stress fractures, the tibia was the most commonly involved bone (49.1% of cases), followed by the tarsals (25.3%) and the metatarsals (8.8%). 1 Bilateral stress fractures occurred in 16.6% of cases. With the increasing emphasis on exercise for the elderly, stress fractures should not be over- looked in this population. Although stress fractures have been described in nearly every bone, they are more common in the weight-bearing bones of the lower extremities. Specific anatomic sites for stress fractures may be associ- ated with individual sports, such as the humerus in throwing sports, the ribs in golf and rowing, the spine in gymnastics, the lower extremities in running activities, and the foot in gymnastics. Pathogenesis Stress fractures result from exces- sive, repetitive, submaximal loads on bones that cause an imbalance be- tween bone resorption and formation. An abrupt increase in the duration, intensity, or frequency of physical activity without adequate periods of rest may lead to an escalation in os- teoclast activity. During periods of intense exercise, bone formation lags behind bone resorption. A reduction in ultimate strength has been demon- strated in bone specimens subjected to hyperphysiologic loading regi- mens, thus rendering the bone sus- ceptible to microfractures. Under continued loading conditions, these microcracks may propagate and coa- lesce into stress fractures. The etiology of stress fractures is multifactorial. The rate of occur- rence depends on the bone compo- sition, vascular supply, surrounding muscle attachments, systemic fac- tors, and type of athletic activity. From a biomechanical standpoint, stress fractures may be the result of muscle fatigue, which leads to the transmission of excessive forces to the underlying bone. Muscles may also contribute to stress injuries by concentrating forces across a local- ized area of bone, thus causing mechanical insults that exceed the stress-bearing capacity of the bone. 2 Dr. Boden is Adjunct Assistant Professor, Uniformed Services University of the Health Sciences, The Orthopaedic Center, Rockville, Md. Mr. Osbahr is Laboratory Researcher, Duke University Medical Center, Durham, NC. Reprint requests: Dr. Boden, The Orthopaedic Center, #201, 9711 Medical Center Drive, Rockville, MD 20850. Copyright 2000 by the American Academy of Orthopaedic Surgeons. Abstract Stress fractures are common overuse injuries seen in athletes and military recruits. The pathogenesis is multifactorial and usually involves repetitive sub- maximal stresses. Intrinsic factors, such as hormonal imbalances, may also con- tribute to the onset of stress fractures, especially in women. The classic presenta- tion is a patient who experiences the insidious onset of pain after an abrupt increase in the duration or intensity of exercise. The diagnosis is primarily clini- cal, but imaging modalities such as plain radiography, scintigraphy, computed tomography, and magnetic resonance imaging may provide confirmation. Most stress fractures are uncomplicated and can be managed by rest and restriction from the precipitating activity. A subset of stress fractures can present a high risk for progression to complete fracture, delayed union, or nonunion. Specific sites for this type of stress fracture are the femoral neck (tension side), the patella, the anterior cortex of the tibia, the medial malleolus, the talus, the tarsal navicu- lar, the fifth metatarsal, and the great toe sesamoids. Tensile forces and the rela- tive avascularity at the site of a stress-induced fracture often lead to poor healing. Therefore, high-risk stress fractures require aggressive treatment. J Am Acad Orthop Surg 2000;8:344-353 High-Risk Stress Fractures: Evaluation and Treatment Barry P. Boden, MD, and Daryl C. Osbahr Barry P. Boden, MD, and Daryl C. Osbahr Vol 8, No 6, November/December 2000 345 In addition to extrinsic mechani- cal influences, systemic factors, such as hormonal imbalances, nutritional deficiencies, sleep deprivation, col- lagen abnormalities, and metabolic bone disorders, may contribute to the development of stress fractures. A high incidence of stress fractures has been reported in women. 3 There- fore, it is especially important to investigate intrinsic factors in female athletes. The term “female athlete triad” refers to the combina- tion of an eating disorder, amenor- rhea, and osteoporosis. Women participating in high-level figure skating, gymnastics, and cross- country running are particularly prone to this triad. In an effort to minimize body fat and maintain high athletic performance, many young women develop eating disor- ders during puberty. Amenorrhea and oligomenorrhea are common findings in competitive female dis- tance runners, with the prevalence of menstrual irregularities as high as 50%. 3 The resultant estrogen- deficient state leads to decreased bone mineral density and an in- creased risk of stress fractures. Male endurance athletes are also predisposed to stress fractures due to abnormally low sex hormone levels. 4 Testosterone levels may decrease by as much as 25% within 2 days of vigorous training. Testos- terone inhibits interleukin-6, a cytokine responsible for enhancing osteoclast development. Therefore, low levels of testosterone in endur- ance athletes result in increased os- teoclast production and bone re- sorption. Clinical Evaluation Early accurate diagnosis is essential for avoiding both complications and a prolonged delay of return to competition. There is commonly an insidious onset of pain over a period of days to weeks. Symptoms are aggravated by activity and relieved with rest. The initial evaluation should include a review of the exer- cise or training program, especially any recent changes in the type or level of activity. The patient’s gen- eral health, medications, diet, and occupation and related activity should also be assessed, as well as the menstrual history in women. Bone tenderness is the most obvi- ous finding on physical examination. In areas that are not superficial, and therefore are not palpable, such as the femoral neck, pain can be elicited by putting the extremity through a gentle range of motion. Superficial stress fractures may exhibit local swelling and palpable periosteal thickening. Evaluation of limb bio- mechanics is essential to identify risk factors such as muscle imbal- ance, limb-length discrepancy, and excessive subtalar pronation. The differential diagnosis for stress fractures includes stress reac- tion, which is visualized as an area of prefracture bone remodeling. Unlike stress fractures, in stress reactions the bone is weakened but not physically disrupted. Other pathologic processes in the differen- tial diagnosis include periostitis, infection, avulsion injuries, muscle strain, bursitis, neoplasm, exertional compartment syndrome, and nerve entrapment. In the pelvis and fe- mur, the diagnosis is often delayed because the lesion is mistaken for a more benign condition, such as muscle strain or bursitis. Imaging Although a stress fracture is usually suspected on the basis of the find- ings from a thorough history and physical examination, specific im- aging modalities may be helpful in confirming the diagnosis or provid- ing more information. Radio- graphs are typically normal for the first 2 to 3 weeks after the onset of symptoms and may reveal no find- ings for several months. Periosteal reaction, cortical lucency, or even a fracture line may be appreciated on later films. Radionuclide imaging has tradi- tionally been considered to be a sensitive method of confirming clinically suspected stress fractures. In the early stages of a stress frac- ture, before changes are visualized on plain films, bone scans are high- ly sensitive for detecting stress in- juries. 5 Acute stress fractures are depicted as discrete, localized, sometimes linear areas of increased uptake on all three phases of a tech- netium-99m diphosphonate bone scan. Soft-tissue injuries are charac- terized by increased uptake only in the first two phases. As healing of the stress fracture occurs, the flow, or angiographic, phase (phase I) may revert to normal, followed by the blood-pool, or soft-tissue imag- ing, phase (phase II). The intensity of activity on delayed images (phase III) decreases over 3 to 18 months as the bone remodels, often lagging behind clinical resolution of symptoms. Therefore, bone scans should not be used to monitor heal- ing and dictate return to activity. Technetium bone scanning is highly sensitive for detecting stress fractures but lacks specificity. In the case of patients with clinically suspected stress-related injuries that are radiographically negative, bone scintigraphy is more sensitive than magnetic resonance (MR) imaging as the initial imaging modality, particularly in evaluating suspected lesions in the spine or pelvis, identifying multiple stress fractures, and distinguishing bipar- tite bones from stress fractures. However, scintigraphy may be overly sensitive; as many as 50% of scintigraphically positive findings can occur at asymptomatic sites. Areas of increased uptake may rep- resent subclinical sites of bone remodeling or stress reactions. High-Risk Stress Fractures Journal of the American Academy of Orthopaedic Surgeons 346 Stress reactions are classified as pre–grade I lesions and are charac- terized by increased uptake in all three phases in the superficial or periosteal layer of bone. The clinical significance of these lesions is con- troversial. Although some progress to stress fractures if untreated, oth- ers resolve despite continued train- ing. 6 A brief rest period is appropri- ate for these lesions. The development of single-photon- emission computed tomography has enhanced the contrast resolution of images by eliminating the surround- ing soft tissue. This has resulted in improved detection and localization of small stress fractures, especially in the spine and pelvis. Magnetic resonance imaging is a valuable tool in identifying stress fractures when the clinical diagnosis is in doubt. Besides having higher specificity than scintigraphy in dis- tinguishing bone involvement from soft-tissue injuries, MR imaging is helpful in grading the stage of cer- tain stress fractures and, therefore, predicting the time to recovery. 7 In addition, MR imaging avoids radia- tion exposure and requires less time than three-phase bone scintig- raphy. Although MR imaging may be slightly more expensive than scintigraphy, the added sensitivity makes the test cost-effective. A classification system for grad- ing stress fractures with scintigraphy and/or MR imaging has been pro- posed (Table 1). 8 By grading stress fractures, the approximate time to healing can be predicted. Indications for obtaining an MR study include a suspicion of a femoral neck stress fracture in an athlete. Treatment Overview The first step in treating stress frac- tures is identifying and correcting any predisposing factors. Analysis of training techniques by an experi- enced coach can be helpful in re- ducing recurrences. Intrinsic fac- tors, such as hormonal, nutritional, and medical abnormalities, also need to be assessed and corrected. Medical evaluation should be con- sidered for any patient with risk fac- tors, such as recurrent or multiple stress fractures and stress fractures with delayed healing. In amenor- rheic or oligomenorrheic athletes, a return to normal menstrual function can often be achieved by decreasing the intensity of the training regimen. Replacement therapy with oral con- traceptives or estrogen can also help hasten the return to normal men- strual cycles, thereby improving bone mineral density. Stress fractures may be broadly classified as either low-risk or high- risk injuries. The differentiation is important in both the diagnostic workup and treatment. Low-risk stress fractures infrequently require expensive imaging modalities (e.g., scintigraphy, CT, or MR imaging) for evaluation. Most cases can be diagnosed on the basis of a thor- ough history, physical examination, and radiographs. If the plain films are normal but the level of clinical suspicion is high, a trial of rest and evaluation with serial radiographs is appropriate. A rest period of 1 to 6 weeks of limited weight bearing progressing to full weight bearing may be necessary. This is followed by a phase of low-impact activities, such as biking and swimming. Table 1 Radiologic Grading System for Stress Fractures * Grade Radiograph Bone Scan MR Imaging † Treatment 1 Normal Mild uptake confined Positive STIR image Rest for 3 weeks to one cortex 2 Normal Moderate activity; Positive STIR and Rest for 3-6 weeks larger lesion confined T2-weighted images to unicortical area 3 Discrete line (+/-), Increased activity No definite cortical Rest for 12-16 weeks periosteal reaction (+/-) (>50% width of bone) break; positive T1- and T2-weighted images 4 Fracture or periosteal More intense Fracture line; positive T1- Rest for 16+ weeks reaction bicortical uptake and T2-weighted images * Adapted with permission from Arendt EA, Griffiths HJ: The use of MR imaging in the assessment and clinical management of stress reactions of bone in high-performance athletes. Clin Sports Med 1997;16:291-306. † STIR = short-tau inversion sequence. Barry P. Boden, MD, and Daryl C. Osbahr Vol 8, No 6, November/December 2000 347 Once the patient can perform low- impact activities for prolonged peri- ods without pain, high-impact exer- cises may be initiated. Typically, the athlete gradually increases jogging mileage and eventually returns to sport-specific activities. High-risk stress fractures have a predilection for progressing to com- plete fracture, delayed union, or nonunion; therefore, they present treatment challenges and require a more aggressive approach. These problematic stress fractures include those in the femoral neck (tension side), patella, anterior cortex of the tibia, medial malleolus, talus, tarsal navicular, fifth metatarsal, and great toe sesamoids. When radiographs are positive, use of other imaging modalities is usually not necessary. In athletes who have chronic pain and normal findings on initial radio- graphs, a bone scan, CT scan, or MR study is recommended. Because of the high complication rate, high-risk stress fractures should be treated like acute fractures. When the diag- nosis is delayed, nonoperative treat- ment is less successful. An algorithm has been formulated to help guide treatment of high-risk stress fractures (Fig. 1). Fractures that are scintigraphically positive yet radiographically negative should be treated with a period of rest. The same treatment is appropriate for low-risk stress fractures. If the stress fracture becomes evident on plain films, treatment should be individu- alized. For most high-risk stress frac- tures of the lower leg and foot, an aggressive nonoperative protocol consisting of non-weight-bearing cast immobilization is recommended, especially if the diagnosis is made soon after the onset of symptoms. The exception to this rule is the tension-side femoral neck stress frac- ture, which requires internal fixation to prevent the potentially devastating complications of fracture progres- sion. In the high-performance athlete whose livelihood is dependent on early return to competition or the athlete who demands an early return to activity, surgical intervention is appropriate. Displaced stress frac- tures and stress fractures with chron- ic radiographic findings, such as intramedullary sclerosis or cystic changes, also require operative inter- vention. Prevention Stress fractures are best managed by prevention. Training errors, such as an excessive increase in intensity, are the most frequent culprit and should be corrected. New activi- ties, such as hill running or run- ning on a hard surface, may be con- tributing factors. The type and condition of the running shoes should also be assessed. For mili- tary personnel, improvement of boots, such as the use of viscoelas- tic insoles, may help reduce the incidence of lower-extremity stress fractures. Athletes, coaches, mili- tary personnel, and parents should be educated about the deleterious effects of overtraining and the im- portance of periodic rest days. In addition, female athletes and their coaches need to be alerted to the adverse effects of eating disorders and hormonal abnormalities. • Radiography • Bone scan, CT, or MR study if site and/or duration of symptoms warrant Surgery Surgery Positive radiograph Negative radiograph but positive bone scan, CT, or MR study Negative radiograph Suspected stress fracture Nonoperative treatment Low-demand athlete Trial of nonoperative treatment High-demand athlete Nondisplaced fracture Displaced fracture or chronic changes If unsuccessful, surgical treatment Figure 1 Algorithm for evaluation and treatment of suspected high-risk stress fractures. High-Risk Stress Fractures Journal of the American Academy of Orthopaedic Surgeons 348 Causation and Treatment of Specific Injuries Femoral Neck Stress fractures of the femoral neck, although uncommon, have a high complication rate if the diag- nosis is missed or the patient is im- properly treated. 9 These fractures may develop as the hip muscu- lature becomes fatigued with pro- longed activity and subsequently loses its protective shock-absorbing effect. Intrinsic factors, such as coxa vara and osteopenia, also may predispose the femoral neck to in- jury. The primary presenting symp- tom is pain in the anterior groin re- gion. The pain is often exacerbated by weight bearing, resulting in an antalgic gait. Bone tenderness is difficult to elicit due to the overly- ing soft tissue. Hip motion, espe- cially at the extremes of internal and external rotation, may be pain- ful or limited. Stress fractures can occur on either the compression or the tension side of the femoral neck. 10 In the more common com- pression stress fractures, the injury begins at the inferior cortex of the femoral neck. As complete dis- placement is extremely rare, nonop- erative treatment is appropriate. The second type, the distraction (or tension) stress fracture, starts in the superior cortex of the femoral neck and may advance across the femoral neck as a fracture line per- pendicular to the axis of the femoral neck. Magnetic resonance imaging is highly sensitive in identifying and delineating the extent of stress frac- tures in the femoral neck. Distraction fractures have a greater propensity to become displaced with continued stress than compression stress frac- tures and, therefore, require more aggressive treatment. Complications, such as delayed union, nonunion, varus deformity, and osteonecrosis, may develop after a displaced frac- ture. In one study, 11 60% of patients with a displaced femoral neck frac- ture that was appropriately treated were unable to return to their prein- jury activity level. Although nonoperative treat- ment of distraction stress fractures of the femoral neck can be success- ful, initial open reduction and in- ternal fixation (ORIF) avoids the morbidity of a potential complete fracture. 12 In patients with widen- ing of the cortical break or displace- ment, cannulated screws should be placed percutaneously on an urgent basis. Displaced stress fractures often require more time to heal than acute traumatic fractures. Therefore, postoperatively, the patient should be maintained in non-weight-bearing status for 6 weeks, followed by an additional 6 weeks of partial weight bearing. Patella Stress fractures of the patella have been reported in athletes and in cerebral palsy patients with knee-flexion contractures. They may also be a complication of total knee replacement. 13 An apparent bipartite patella that is sympto- matic and is visualized as increased uptake on nuclear scanning may actually be an acute or chronic stress fracture. In athletes, stress fractures may occur in either a longitudinal or a transverse direction. It has been postulated that transverse stress fractures initiate on the anterior sur- face of the patella due to repeated tension forces from the quadriceps and patellar tendons with the knee in flexion. 13 During stance, the quad- riceps force required to stabilize the knee is greater than 200% of body weight. Transverse patellar stress frac- tures have also been identified after anterior cruciate ligament recon- struction. The cause may be related to stress risers created at the bone- plug defect after harvesting of a bone-patella-bone graft. In addi- tion, any postoperative knee-flexion contracture may result in increased patellofemoral forces predisposing to a stress fracture. Stress reactions in the patella that are scintigraphi- cally positive but are not evident on plain films may be an underappre- ciated clinical entity causing patel- lofemoral pain after reconstruction of the anterior cruciate ligament. Untreated or misdiagnosed patellar stress fractures may develop into complete fractures during athletic activity. Because the overall incidence of patellar stress fractures is low and the number of reported series is small, it is difficult to make defini- tive treatment recommendations. Nonetheless, restriction of activity is appropriate when nondisplaced fractures are identified on scintig- raphy and are not seen on radio- graphs. Once the stress fracture is evident on radiographs, treatment should be individualized to the patient. Nonoperative treatment with careful observation is recom- mended for the patient who does not require an immediate return to activity. For high-demand athletes and those with displaced fractures and nonunions, ORIF is appropriate. A standard tension-band wiring technique with Kirschner wires or cannulated compression screws pro- vides excellent fixation. Tibia In athletes, the tibial shaft is the most common location of stress fractures. 1 Depending on the pa- tient population, the incidence may range from 20% to 75% of all stress fractures. Tibial stress fractures may occur at any site along the shaft of the bone, but are most frequently encountered in the posteromedial cortex (compression side). The vast majority of tibial stress fractures are transverse in orientation, but longi- tudinal stress fractures have also been reported. Longitudinal stress fractures often have an atypical pre- Barry P. Boden, MD, and Daryl C. Osbahr Vol 8, No 6, November/December 2000 349 sentation, necessitating MR imaging for definitive diagnosis. Tibial stress fractures on the posteromedial cor- tex respond favorably to discontinu- ation of the inciting activity. Supplemental use of a pneumatic brace may allow athletes to return to activity sooner than traditional treatment alone. 14 In a randomized, prospective study, the 10 patients treated with a brace returned to full unrestricted activity at an average of 21 days, compared with 77 days for the 8 patients treated with rest alone. The pneumatic brace may function as a partial-weight-bearing cast or as a venous tourniquet that shifts electrolytes into the interstitial fluid space to stimulate osteoblastic bone formation. A less common, but more prob- lematic, location of tibial stress frac- tures is the anterior cortex of the middle third of the tibia (Fig. 2). Findings on full-length tibial radio- graphs are subtle, often leading to a delay in diagnosis. Therefore, careful scrutiny of the plain films with a magnifying glass, especially the lateral view, is essential when a patient presents with pain in the middle third of the tibia. Radio- graphs focusing on the middle third may reveal an anterior tibial stress fracture. Both constant tension from pos- terior muscle forces and hypovas- cularity of the anterior aspect of the tibia predispose this site to non- union or delayed union. Histo- pathologic examination of chronic anterior-cortex tibial stress fractures has revealed fibrotic infiltrations, local osteonecrosis, and limited or no bone-remodeling activity, con- sistent with pseudarthrosis. In contrast to compression tibial stress fractures, which usually occur in distance runners, tension tibial stress injuries typically occur in athletes performing repetitive jumping and leaping activities. Patients present with point tender- ness over the anterior aspect of the central third of the tibia. These fractures have the potential to progress to complete fractures. 15 Radiographs are often initially nor- mal, but subsequently develop a characteristic V-shaped (or wedge) defect in the anterior cortex, with the open end of the “V” being directed anteriorly. Callus forma- tion is generally absent. 16 The radio- graphic appearance has also been referred to as the “dreaded black line” because of the prolonged heal- ing time. Once the cortex becomes hypertrophied and the fissure widens, the healing capacity is extremely limited. Radionuclide imaging may demonstrate minimal activity, indicating nonunion of the stress fracture. The differential diagnosis should include infection, tumors, medial tibial stress syn- drome, and exertional compart- ment syndrome. Initial treatment of stress frac- tures of the anterior midportion of the tibia is generally a trial of rest, with or without immobilization, for a minimum of 4 to 6 months. If the radiographs reveal chronic changes, such as a wide fissure, surgical intervention becomes necessary. Numerous treatments have been proposed for stress fractures that display delayed union. Prompt healing has been reported after exci- sion and bone grafting of the le- sion. 15 In a series in which a regi- men of rest and external electrical stimulation was evaluated, 16 seven of eight patients showed complete healing after an average of 8.7 months of treatment. Other authors have described less favorable results with electromagnetic stimulation. 15 Chang and Harris 17 reported good to excellent results in five patients with recalcitrant stress fractures that were treated with reamed unlocked tibial nails. Intramedullary fixation has become the favored approach for recalcitrant anterior-cortex tibial stress fractures. Medial Malleolus The medial malleolus is a rela- tively uncommon site for stress fractures, but they can occur in ath- letes participating in running and jumping activities. Repetitive im- pingement of the talus on the medial malleolus during ankle dorsiflexion and tibial rotation may result in a medial malleolar stress fracture. Patients with medial malleolar stress fractures present with tender- ness over the medial malleolus and an ankle effusion. They may have pain during athletic activities for several weeks prior to an acute epi- sode. The pain increases with activ- ity and is relieved by rest. The frac- ture line is vertical or oblique and originates from the junction of the tibial plafond and the medial malle- olus. 18 For individuals with negative radiographs and a positive bone scan or an incomplete fracture visu- alized on MR imaging, treatment is individualized on the basis of the level of athletic activity. Most pa- Figure 2 Lateral radiograph of the tibia demonstrates delayed union of an anterior- cortex tibial stress fracture (arrow). Peri- osteal thickening surrounds the cortical break. High-Risk Stress Fractures Journal of the American Academy of Orthopaedic Surgeons 350 tients are successfully treated with cast immobilization or ankle brac- ing and avoidance of impact activi- ties. Athletes desiring early return to competition may be treated with percutaneous drilling and immobi- lization or internal fixation. 19 Both surgical and nonsurgical treatment usually result in a full return to activity; however, resolution of symptoms may take 4 to 5 months with nonoperative therapy. 20 Inter- nal fixation with malleolar screws is advocated for patients with a com- plete fracture line on radiographs. 18 Due to the high shear forces exerted at the fracture site, nonunion may develop. 20 In this circumstance, ORIF with two cancellous screws is required. Bone grafting is indicated when there is fracture displacement and, in chronic cases, when sclerosis is present at the fracture site. Talus Stress fractures of the lateral process of the talus are extremely rare. 21 Patients present with lateral ankle or sinus tarsi pain that is exac- erbated by activity. Excessive sub- talar pronation and plantar flexion may predispose athletes to injury as the lateral process of the calcaneus impinges on the concave posterolat- eral corner of the talus. Alterna- tively, a supinated foot may concen- trate forces on the lateral process of the talus. Plain films often fail to re- veal the stress fracture. Computed tomographic scans are helpful in identifying the lesion at the postero- lateral border of the talus. The stress fracture often extends into the subta- lar joint, which explains the symp- toms in the region of the sinus tarsi. Outcomes after early return to activ- ity are poor. Therefore, a 6-week trial of non-weight-bearing cast im- mobilization is recommended, fol- lowed by rehabilitation and use of an orthosis to correct any excess pro- nation. Tarsal Navicular Tarsal navicular stress fractures occur primarily in active athletes involved in sprinting and jumping sports. 22,23 The presentation typi- cally involves an insidious onset of nondescript pain in the medial arch area that is aggravated by activity. Findings on examination are usu- ally limited to tenderness over the navicular with occasional limitation of subtalar motion or dorsiflexion of the ankle. Navicular stress fractures occur in the sagittal plane in the central third of the bone or at the junction of the central and lateral thirds of the navicular. 24 This site corresponds to the zone of maxi- mum shear stress on the navicular from the surrounding bones. The lesion begins at the proximal dorsal articular surface and propagates in a distal and plantar direction, result- ing in a partial or complete injury. Microangiographic studies have demonstrated that the navicular is supplied by peripheral, medial, and lateral vessels, leaving the central third relatively avascular. 22 The diagnosis of navicular stress fracture is often missed on routine radiographs because the tarsal na- vicular lies in an oblique direction (Fig. 3). 22 When the diagnosis is sus- pected, an anatomic anteroposterior radiograph should be obtained with the foot inverted. Additional radio- Figure 3 A, Standard anteroposterior radiograph of a professional basketball player with midfoot pain does not reveal a stress fracture. B, Bone scan reveals increased uptake in the navicular bone. C, CT scan demonstrates a complete navicular stress fracture. A B C Barry P. Boden, MD, and Daryl C. Osbahr Vol 8, No 6, November/December 2000 351 logic studies may include bone scan- ning, tomography, CT, or MR imag- ing. Bone scans are sensitive for detecting navicular stress fractures, but are not specific for distinguish- ing stress reaction from stress frac- ture. Computed tomography with reconstructions and MR imaging are more sensitive than bone scanning and provide information on the ex- tent of the lesion. Patients who have an early diag- nosis of partial or complete navicular stress fracture have a high union rate if treated for 6 to 8 weeks with non- weight-bearing cast immobiliza- tion. 22 When weight bearing is per- mitted initially, the risk of delayed union, nonunion, or recurrence is dramatically higher. Minimally dis- placed navicular stress fractures may be treated with cast immobilization or ORIF. Displaced fractures, de- layed unions, and nonunions are best treated with ORIF and bone grafting. Fixation is accomplished by means of one or two compression screws placed across the fracture. 24 A semirigid molded orthosis is rec- ommended for arch support during the rehabilitation phase and after return to athletic activity. Fifth Metatarsal Stress fractures of the fifth meta- tarsal occur at the proximal diaph- ysis of the bone just distal to the tuberosity and the ligamentous structures. This injury has a high incidence in basketball players. Fifth-metatarsal stress fractures have a propensity for delayed union or nonunion and have a high risk of refracture after nonoperative treat- ment. 25 An acute injury is often pre- ceded by a 2- to 3-week history of discomfort over the lateral aspect of the foot. On examination, point ten- derness is present at the site of the stress fracture. Pain is exacerbated by inversion of the foot. Radio- graphs reveal a radiolucent line with variable degrees of periosteal re- action and intramedullary sclerosis. Treatment depends on the stage of the lesion. For the patient with prodromal symptoms but negative radiographs, avoidance of weight- bearing activity and a semirigid metatarsal functional brace that unloads the fifth metatarsal can be curative. 26 If symptoms persist for more than 3 weeks or if radiographs reveal a stress fracture, treatment options include non-weight-bearing cast immobilization for 6 weeks or intramedullary-screw fixation. For high-demand athletes, internal fixa- tion with a compression screw pro- vides good results and faster return to activity. 27 For patients with a de- layed union and medullary sclerosis on radiographs, intramedullary fixa- tion with curettage is recommended. To avoid reinjury, a functional me- tatarsal brace should be used for at least 1 month after surgery. Careful preoperative assessment of the radiographs to determine the width and length of the screw is critical to avoid intraoperative complications, such as iatrogenic fracture of the metatarsal. In an average-size adult, a cannulated 4.5-mm lag screw is preferred. Drilling the intramedullary canal before screw placement to debride the intramedullary fibrous tissue is recommended. The screw should be countersunk to avoid skin irrita- tion at the base of the bone. Axial screw fixation is performed with- out opening the fracture site and risking further damage to the blood supply. Full return to competition can usually be achieved by 8 to 10 weeks. Great Toe Sesamoids The function of the great toe sesamoids is to diminish the pres- sure on the metatarsal head and to provide a mechanical advantage to the flexor hallucis brevis. Sesa- moid stress fractures (Fig. 4) are uncommon, but may lead to pro- longed disability if misdiagnosed or treated incorrectly. 28 The injury has a slight predominance at the medial sesamoid, which lies directly under the head of the first meta- tarsal. Repeated dorsiflexion of the great toe during running and jump- ing can result in tensile forces on the sesamoid sufficient to cause a transverse stress fracture. The clin- ical diagnosis is suggested by ten- derness directly over the plantar Figure 4 Anteroposterior (A) and lateral (B) radiographs of the foot of a soccer player show a medial sesamoid stress fracture. A B High-Risk Stress Fractures Journal of the American Academy of Orthopaedic Surgeons 352 aspect of the first metatarsopha- langeal joint, discomfort with maxi- mum dorsiflexion of the first toe, or push-off disability. Radiographs should include weight-bearing anteroposterior and lateral views as well as an axial view centered on the sesamoids. Serial images may be helpful in dis- tinguishing a stress fracture from an acute fracture. The reported incidence of bipartite sesamoid in the general population ranges from 5% to 30%, and the incidence of bilaterality is approximately 80%. Most bipartite sesamoids occur in the medial sesamoid. Radiograph- ically, the bipartite sesamoid has smooth edges and is larger than the undivided sesamoid. Findings suggestive of a stress fracture include a transverse fracture line with jagged margins. Nuclear scanning can help differentiate an acute fracture or a stress fracture from a bipartite sesamoid. Acute stress fractures are treated for 6 weeks with a non-weight- bearing cast that extends to the dis- tal tip of the toe to prevent dorsi- flexion. Because the incidence of delayed union, nonunion, or recur- rence is so high, early surgical inter- vention is appropriate for selected patients. In chronic cases, histo- logic specimens have demonstrated fibrous nonunion or pseudarthrosis with large resorptive cavities 28 ; therefore, the threshold for opera- tive treatment is lowered. Surgery involves complete ex- cision of the sesamoid with careful dissection to avoid disruption of the flexor hallucis brevis. Bone grafting without excision also provides satis- factory results. In the rare case of concomitant medial and lateral sesamoid stress fractures, partial sesamoidectomy is preferred over complete excision to avoid a cock- up deformity. Postoperatively, a cast is applied for 2 to 3 weeks, fol- lowed by gradual resumption of activities. Summary Stress fractures are caused by repet- itive submaximal forces that exceed the adaptive ability of the bone. Training errors are the most com- mon precipitating factors; however, intrinsic systemic factors should not be overlooked. Stress fractures are classified as low-risk or high-risk injuries. Low-risk stress fractures infrequently require expensive imaging modalities for diagnosis and generally respond to activity modification. High-risk stress frac- tures have a propensity to develop into chronic injuries; therefore, more aggressive treatment is neces- sary. Sophisticated imaging studies are often necessary when the diag- nosis is in doubt or when the extent of the injury is difficult to establish. References 1. Matheson GO, Clement DB, McKenzie DC, Taunton JE, Lloyd-Smith DR, MacIntyre JG: Stress fractures in ath- letes: A study of 320 cases. Am J Sports Med 1987;15:46-58. 2. Stanitski CL, McMaster JH, Scranton PE: On the nature of stress fractures. Am J Sports Med 1978;6:391-396. 3. Barrow GW, Saha S: Menstrual irreg- ularity and stress fractures in colle- giate female distance runners. Am J Sports Med 1988;16:209-216. 4. Voss LA, Fadale PD, Hulstyn MJ: Exercise-induced loss of bone density in athletes. J Am Acad Orthop Surg 1998;6:349-357. 5. Prather JL, Nusynowitz ML, Snowdy HA, Hughes AD, McCartney WH, Bagg RJ: Scintigraphic findings in stress fractures. J Bone Joint Surg Am 1977;59:869-874. 6. Chisin R, Milgrom C, Giladi M, Stein M, Margulies J, Kashtan H: Clinical significance of nonfocal scintigraphic findings in suspected tibial stress frac- tures. Clin Orthop 1987;220:200-205. 7. Fredericson M, Bergman AG, Hoffman KL, Dillingham MS: Tibial stress reac- tion in runners: Correlation of clinical symptoms and scintigraphy with a new magnetic resonance imaging grading system. Am J Sports Med 1995; 23:472-481. 8. Arendt EA, Griffiths HJ: The use of MR imaging in the assessment and clinical management of stress reactions of bone in high-performance athletes. Clin Sports Med 1997;16:291-306. 9. Shin AY, Gillingham BL: Fatigue frac- tures of the femoral neck in athletes. J Am Acad Orthop Surg 1997;5:293-302. 10. Devas MB: Stress fractures of the fe- moral neck. J Bone Joint Surg Br 1965; 47:728-738. 11. Johansson C, Ekenman I, Törnkvist H, Eriksson E: Stress fractures of the femoral neck in athletes: The conse- quence of a delay in diagnosis. Am J Sports Med 1990;18:524-528. 12. Aro H, Dahlstrom S: Conservative management of distraction-type stress fractures of the femoral neck. J Bone Joint Surg Br 1986;68:65-67. 13. Teitz CC, Harrington RM: Patellar stress fracture. Am J Sports Med 1992; 20:761-765. 14. Swenson EJ Jr, DeHaven KE, Sebastia- nelli WJ, Hanks G, Kalenak A, Lynch JM: The effect of a pneumatic leg brace on return to play in athletes with tibial stress fractures. Am J Sports Med 1997;25:322-328. 15. Green NE, Rogers RA, Lipscomb AB: Nonunions of stress fractures of the tibia. Am J Sports Med 1985;13:171-176. 16. Rettig AC, Shelbourne KD, McCarroll JR, Bisesi M, Watts J: The natural his- tory and treatment of delayed union stress fractures of the anterior cortex of the tibia. Am J Sports Med 1988;16: 250-255. 17. Chang PS, Harris RM: Intramedullary nailing for chronic tibial stress frac- tures: A review of five cases. Am J Sports Med 1996;24:688-692. 18. Shelbourne KD, Fisher DA, Rettig AC, McCarroll JR: Stress fractures of the medial malleolus. Am J Sports Med 1988;16:60-63. 19. Orava S, Karpakka J, Taimela S, Hulkko A, Permi J, Kujala U: Stress Barry P. Boden, MD, and Daryl C. Osbahr Vol 8, No 6, November/December 2000 353 fracture of the medial malleolus. J Bone Joint Surg Am 1995;77:362-365. 20. Reider B, Falconiero R, Yurkofsky J: Nonunion of a medial malleolus stress fracture: A case report. Am J Sports Med 1993;21:478-481. 21. Bradshaw C, Khan K, Brukner P: Stress fracture of the body of the talus in ath- letes demonstrated with computer to- mography. Clin J Sport Med 1996;6:48-51. 22. Torg JS, Pavlov H, Cooley LH, et al: Stress fractures of the tarsal navicular: A retrospective review of twenty-one cases. J Bone Joint Surg Am 1982;64:700-712. 23. Khan KM, Fuller PJ, Brukner PD, Kearney C, Burry HC: Outcome of conservative and surgical management of navicular stress fracture in athletes: Eighty-six cases proven with comput- erized tomography. Am J Sports Med 1992;20:657-666. 24. Fitch KD, Blackwell JB, Gilmour WN: Operation for non-union of stress frac- ture of the tarsal navicular. J Bone Joint Surg Br 1989;71:105-110. 25. Torg JS, Balduini FC, Zelko RR, Pavlov H, Peff TC, Das M: Fractures of the base of the fifth metatarsal distal to the tuber- osity: Classification and guidelines for non-surgical and surgical management. J Bone Joint Surg Am 1984;66:209-214. 26. Dameron TB Jr: Fractures of the proxi- mal fifth metatarsal: Selecting the best treatment option. J Am Acad Orthop Surg 1995;3:110-114. 27. DeLee JC, Evans JP, Julian J: Stress fracture of the fifth metatarsal. Am J Sports Med 1983;11:349-353. 28. Van Hal ME, Keene JS, Lange TA, Clancy WG Jr: Stress fractures of the great toe sesamoids. Am J Sports Med 1982;10:122-128.