Management and Prevention of Bone Stress Injuries in Long-Distance Runners

Factors Modifying the Load Applied to a Bone

The load applied to a bone during running represents the summation of external and internal forces, the components of which are magnitude, rate, frequency, duration, and direction. These components influence, respectively, the magnitude, rate, frequency, duration, and location of bone strain. Unfortunately, it is not currently possible to precisely measure the total load being applied to a bone, although subject-specific musculoskeletal models are being developed for this purpose. In the absence of an available measure of total bone loading during running, surrogate measures have been utilized, including measuring (1) bone strains by invasively attaching a strain gauge to a bone of interest, (2) bone acceleration (often referred to as shock) using an accelerometer, and (3) ground reaction forces (GRFs) obtained with a force platform. While each of these methods has limitations in defining the total load being applied to a bone, they have provided insight into the factors that influence BSI risk.

Biomechanical Factors Faulty biomechanics can contribute to BSI risk and can be divided into those related to abnormal forces and those related to abnormal motions. Increased forces on a normally aligned lower extremity can result in abnormal bone loading. Alternatively, normal forces applied to a malaligned lower limb can also abnormally load the skeleton. Having the combination of abnormal forces coupled with a malaligned lower limb is thought to further amplify BSI risk.

Runners with abnormal loading (high GRF magnitudes and rates, and accelerations during the early stance phase of running gait) are proposed to be at heightened risk of BSI. While variable evidence supports the magnitude of GRF impact peak as a discriminator between those with and without a BSI history,^{10,29,54,86,87} individuals with a BSI history have greater GRF loading rates and peak accelerations.^87,100 Similarly, a preliminary prospective study by Davis et al³² suggested that subjects who sustained a BSI had higher peak acceleration and vertical GRF loading rates prior to injury than matched controls. Torsional loads have also been associated with BSI history. Milner et al⁸⁶ reported significantly higher free moments in runners with a history of tibial BSI.

Abnormal movement patterns can also increase the risk of BSI. Static alignment is likely to influence movement patterns and has been implicated in BSI development. For example, Williams et al^122,123 demonstrated that high-arched individuals exhibited reduced joint excursions and higher stiffness, and had a greater history of BSI. Other static variables reported to be related to BSI include increased external rotation range of motion of the hip,⁴⁴ leg-length discrepancy,¹² pes planus,¹¹¹ and pes cavus.¹⁰⁷ Individuals with a history of a BSI have been shown to have greater peak hip adduction, knee internal rotation, and peak rearfoot eversion in the frontal plane during running,^37,88,100 and may have less knee flexion in the sagittal plane.⁸⁷ These movement patterns have the potential to alter the magnitude and/or rate of bone loading. They may also alter the direction in which a bone is loaded and the subsequent distribution of strain within the bone. The net result may be increased loading of a less-accustomed bone site.

Training Factors Chronic introduction of high absolute load magnitudes, rates, and accelerations may reduce bone fatigue life, particularly when the number of loading cycles is high (ie, running long distances). However, the influence of these variables may be most prominent when runners attempt to progress their training. Increases in running speed increase GRFs and rates of introduction during heel-toe running gait,⁵⁵ whereas increases in the duration and/or frequency of running sessions increase the total number of bone-loading cycles. In the absence of a change in the load-bearing capacity of a bone, altered loading associated with large changes in training may contribute to microdamage accumulation and the generation of a BSI. There is evidence from military studies that individuals exposed to large changes in physical activity have heightened risk of BSI. For instance, recruits with a lesser history of regular physical activity prior to the commencement of standardized basic training (ie, those with larger changes in physical activity levels) are at a greater risk of developing BSI.^{27,50,72,84,106} Although most athletes do not introduce changes in their running program and consequent bone-loading environment as extreme as those of military recruits, change is a frequent and required means of adaptation to achieve personal and competitive goals. Incrementing a running program too rapidly or frequently relative to a runner's usual activities is thought to be central to disrupting the balance between bone microdamage formation and removal.

Muscle Factors Training changes may independently contribute to BSI development, but the relative risk associated with the change may be compounded by muscle factors. An intimate mechanical relationship exists between muscle and bone, and it is hypothesized that muscle is protective rather than causative of BSI. During impact loading, muscle is believed to act as an active shock attenuator, helping to reduce loads as they are transmitted proximally along the kinetic chain. When muscles are dysfunctional (weakened, fatigued, or altered in their activation patterns), their ability to attenuate loads becomes compromised, potentially leading to increased loading on the skeleton. For instance, fatigue in laboratory studies caused a decrease in shock attenuation,^80,89 an increase in GRF loading rates and peak accelerations,^25,90 and an increase in bone strain magnitude and rate.^47,83 In addition, fatigue can lead to altered kinematics,³⁴ which may alter the direction in which a bone is loaded, resulting in increased bone strain at less-accustomed bone sites.¹²⁴ Further supporting the protective role of muscle in BSI development are prospective clinical studies demonstrating that BSI susceptibility is directly related to muscle size (girth and cross-sectional area)^4,9,12 and strength.⁵⁹

Running Surface Bone loading and, subsequently, BSI risk may also be influenced by external factors such as running surface. Running surface has historically been considered a contributor to BSI risk. Harder surfaces (eg, asphalt/bitumen) have been hypothesized to increase loading compared to softer surfaces (eg, grass, rubber, and sand). However, the interaction between running surface and injury risk is complex. Runners alter their leg stiffness when running on surfaces of differing compliance, apparently to maintain a constant vertical excursion of their center of mass.⁴³ Leg stiffness decreases when running on stiffer surfaces to normalize, to some extent, GRF magnitude; however, GRF loading rates do appear to be systematically increased when running on surfaces that are less compliant.³⁶ Whether the increased loading rate associated with running on less-compliant surfaces contributes to BSI risk remains unclear, as large epidemiological studies of running injuries have failed to show an association between injuries and training surface after controlling for weekly running distance.^76,114 Ultimately, what may be important with regard to BSI risk is whether there has been a recent change in running surface to which the runner has yet to become accommodated. Changes may include increased running on (1) less-compliant surfaces (eg, changing from treadmill to overground running), which may increase bone strain magnitudes and rates⁸²; (2) very compliant surfaces (eg, sand), which may increase energy expenditure and influence muscle-related risk factors and kinematics^97,98; (3) downhill slopes, which may decrease shock attenuation⁹¹ and increase loading magnitudes and rates⁵²; and (4) altered terrain, which may alter kinematics to load less-accustomed skeletal sites.¹²⁵

Shoes and Inserts The role of shoes and inserts (orthoses and insoles) on bone loading and BSI risk is a topic of ongoing debate. Located at the foot-ground interface, shoes and inserts act as filters that theoretically attenuate ground impact forces. In addition, they have the potential to influence motion of the foot and ankle and the subsequent mechanics proximally in the kinetic chain. Via these 2 mechanisms, shoes and inserts may influence bone loading and have an effect on BSI risk. In support, Gardner et al⁵⁰ found that military recruits who trained in shoes of advanced age (an indicator of possible reduced shock-absorptive capacity) were at a greater risk of developing a BSI, and a recent systematic review concluded that orthoses reduced BSI risk during military training.¹⁰⁹ Whether the same benefits are observed in runners remains unclear, and recent work exploring the virtues of barefoot running has raised questions regarding the role of shoes in general injury prevention.⁷⁵

Factors Influencing the Ability of the Bone to Resist Load Without Damage Accumulation

The amount and rate of strain when a load is applied to a bone depend on features of the applied load and also the ability of the bone to resist deformation in the direction of loading. For a given applied load, less-rigid bones experience greater strain at a faster rate than more-rigid bones and are therefore more susceptible to microdamage and BSI formation. Skeletal features that influence bone rigidity include the amount of bone material present (mass) and its distribution (structure), and there is strong evidence that both contribute to BSI risk. For instance, prospective studies have confirmed that BSI susceptibility is inversely related to bone mass and cross-sectional size.^8,9,12,27,51 While it is clear that reduced bone mass and size are risk factors for BSI, it is important to consider modifiable factors that may contribute to these skeletal characteristics. Three modifiable factors in runners that may impact the ability of bone to resist loading and contribute to BSI risk are physical activity history, energy availability, and calcium and vitamin D status.

Physical Activity History A longer history of physical activity appears to be protective against the development of a BSI.^{27,50,72,84,106} An improved ability of the skeleton to resist loading likely contributes to the reduced BSI risk in individuals with a prior history of physical activity. The skeleton responds and adapts to mechanical loading in a site-specific manner to increase its rigidity in the direction of loading. It principally does this during the growing years by preferentially depositing small amounts of new bone on the outer periosteal surface at a distance from the bending axes. As the rigidity of a unit area of bone is proportional to the fourth power of its distance from a bending axis, the addition of a small amount of mass to the outer surface of a bone results in a disproportionate increase in bone strength. The net result is a decrease in the bone strain engendered in response to a given load and an increase in bone fatigue life. For instance, Warden et al¹¹⁹ demonstrated using an animal model that a relatively moderate (less than 10%) increase in bone mass induced by mechanical loading resulted in a large (107-fold) increase in bone fatigue resistance due to the mass being distributed at a distance from the bending axis and reducing bone strain.

Energy Availability Gender factors contribute to BSI susceptibility, with females being at greater risk.¹²¹ The cause for the higher incidence of BSIs in females appears to relate to the interrelationships between energy availability, menstrual function, and bone mass—otherwise known as the female athlete triad. While runners can have one or more components of the triad, low energy availability appears to be the central factor.⁹⁴ Low energy availability results from insufficient dietary intake to meet exercise energy expenditure. It can result from low dietary energy intake (whether inadvertent, intentional, or psychopathological) and/or excessive exercise energy expenditure. While complete discussion of the female athlete triad is beyond the scope of this commentary and has been covered comprehensively elsewhere,⁹⁴ the menstrual and skeletal changes associated with low energy availability reduce the ability of bone to resist load and/or impair its ability to repair microdamage. The net result is heightened BSI risk, particularly in elite female long-distance runners, where the difference between dietary energy intake and exercise energy expenditure is often small and the incidence of menstrual dysfunction is elevated.

Calcium and Vitamin D Status Calcium combines with phosphate to form hydroxyapatite crystals to endow bone with rigidity, whereas vitamin D contributes by promoting calcium absorption in the gut and reabsorption in the kidneys. Prospective studies provide evidence for roles of both low calcium and vitamin D in BSIs. Studying 18- to 26-year-old female distance runners, Nieves and colleagues⁹⁵ demonstrated that women who consumed less than 800 mg of calcium per day had nearly 6 times the BSI rate of those who consumed more than 1500 mg. In terms of vitamin D, Ruohola et al¹⁰⁴ found that male Finnish military recruits with a serum 25-hydroxyvitamin D concentration below a median level of 30.4 ng/mL had a greater risk for BSI. Additional support for the role of both calcium and vitamin D in mediating BSI risk was provided by a randomized controlled trial that demonstrated a 20% reduction in BSI incidence in female Navy recruits with suboptimal baseline daily calcium intake (300 mg) when they were supplemented with 2000 mg of calcium and 800 IU of vitamin D per day.⁷¹ Overall, these data suggest that running athletes should ensure sufficient calcium and vitamin D intakes to meet or exceed the currently recommended dietary allowances of 1000 to 1300 mg and 600 IU (for individuals aged 14–50 years), respectively.⁶⁰

Phase 1: Initial Management

Activity Modification There is no question that temporary discontinuation of running and the introduction of a variable period of modified activity are required in the initial management of low-risk BSIs to permit tissue-level healing and to prevent pathology progression. However, the duration and extent of activity modification are highly variable and decided on an individual basis, using pain as the principal guiding variable. The presence of pain either during or after an activity indicates that the pathological site is being excessively loaded for the current stage of healing, and that loading needs to be titrated.

In the initial stages of BSI management, the goal of the athlete is to be pain free during and after usual activities of daily living (ADLs). Cushioned shoes and/or insoles may provide assistance in dissipating impact forces during ADLs in athletes with a rearfoot or leg BSI,^41,81 whereas stiff-soled shoes may be considered to reduce bending forces and symptoms in athletes with a BSI in the midfoot or forefoot.⁵ Walking should be minimized to that essential to perform ADLs and limited to a normal gait pattern. If a normal gait pattern cannot be used or symptoms are produced either during or after walking, partial weight bearing using assistive gait devices (eg, double or single crutch, cane) should be considered. Alternatively, a pneumatic leg brace may be introduced to promote pain-free gait for BSIs of the fibula or posteromedial border of the tibia.¹¹² If a pain-free normal gait pattern cannot be achieved, a short period in a walking boot or of non-weight bearing may be considered. However, in each scenario, progression to unassisted pain-free gait should be sought as soon as possible.

Athletes need to be pain free not only during but also after activities. The presence of resting and/or night pain is a sign that the underlying pathology may have an inflammatory component. While resting pain does not typically require specific intervention, as it is usually short term and abates with activity modification, some athletes may consider nonsteroidal anti-inflammatory drugs or other analgesics. The use of these agents should be discouraged beyond a few days because of their ability to mask pain and subsequently influence activity progression, and because of the less favorable risk-to-benefit ratio of nonsteroidal anti-inflammatory drugs.¹¹⁶ In particular, nonsteroidal anti-inflammatory drugs have the potential to impede tissue-level healing of BSIs when taken for prolonged periods.^67,74

Identification and Initial Management of Potential Risk Factors The initial period following BSI diagnosis is a useful time to evaluate and begin addressing potential contributing factors, as it is often the time when an athlete gives the most attention. A detailed running and physical activity history is important. Consider not only usual and recent changes in running frequency, duration, and intensity, but also usual or new participation in physical activities beyond running. Bone stress injury risk reflects the sum of all bone loading, and loading from nonrunning activities may be enough to push an athlete beyond his or her injury threshold. Note should also be taken of any recent changes in running surfaces, shoes/inserts, or technique. By combining knowledge of recent running progressions with knowledge of lifelong physical activity history, it may be possible to provide a runner returning from a BSI with advice regarding future running-program design. For instance, novice runners with a minimal physical activity history may need to progress their training program at a slower rate to avoid overloading the skeleton and disrupting the homeostasis between microdamage formation and removal.

History taking in an injured runner should also assess factors influencing the ability of the runner's skeleton to resist loading. It is essential to obtain a history of the patient's general health, medications (including use of glucocorticoids and anticonvulsants), and personal habits to ensure that there are no factors that may influence bone health. A past history of BSI (and other bone injury) and current body mass index of less than 19 kg/m² are strong risk factors for BSI that require assessment and thorough exploration.^64,113 A full dietary history should be taken, with particular attention paid to possible deficits in energy intake and/or eating disorders, and calcium and vitamin D intake. There are numerous online calculators for estimating calcium and caloric intake, with exercising athletes requiring higher-than-recommended intakes for skeletal and body composition maintenance. A maternal family history of osteoporosis (low-trauma fractures) or low bone mineral density should be explored, and a detailed menstrual history should be taken in female runners, including their age of menarche and subsequent menstrual status. Identification of any issues of concern warrants appropriate referral when indicated, remembering that consultation for a BSI may be the first time that a runner's issues associated with the female athlete triad are identified.

In addition to history taking, initial assessment and management of potential factors that influence bone loading during running can be explored, even though running gait cannot typically be assessed until a later time. To guide the interventions and formulate an initial biomechanical impression of the athlete, it is important to gather as much circumstantial evidence as possible. Such evidence may come from recent videos of the athlete running; assessment of shoe wear pattern; a history of overuse injuries; a unilateral predominance; static posture and alignment; and hip, knee, and foot mechanics during nonpainful activities such as walking or single-leg squat. Depending on the individual athlete and the preliminary examination findings, initial biomechanical interventions can be introduced to maintain what the athlete already has or address suspected deficits. Activities can include muscle strength, endurance and control training, and interventions for muscle length and joint mobility. Key areas to consider include control, endurance, and strength at the hip, knee, and ankle; core stability; and strength of the intrinsic and extrinsic muscles of the foot. The activities may need to be modified to be pain free, and are often performed in non-weight bearing or partial weight bearing.

Maintenance of Physical Conditioning Maintaining conditioning during recovery from a BSI is important for a seamless return to running, as the athlete's pathology permits. Conditioning activities should be introduced early, as endurance-trained athletes experience declines in cardiovascular performance in as little as 2 weeks following training cessation.²⁸ There are multiple methods of maintaining cardiovascular fitness during recovery from a BSI, including cycling, swimming, deep-water running (DWR), and antigravity treadmill training (ATT). The latter 2 methods may be most specific to runners, as they more closely reproduce the neuromuscular recruitment patterns involved in running. Antigravity treadmill training is discussed in a later section, as it is not introduced until a runner is pain free during walking and ADLs. In the interim, DWR can be introduced as long as the athlete is pain free both during and after DWR sessions.

Deep-water running is performed in the deep end of a swimming pool, and makes use of buoyancy to provide 100% body-weight support. The runner mimics running on dry land without contacting the floor of the pool, with limb movements being resisted by the viscosity of the water and drag forces. Literature reviews have concluded that DWR introduces cardiovascular demands that are relatively close to those of real running when training at easy to moderate intensities, but that it is less effective at high intensities.^68,101 Thus, DWR is useful in maintaining and not necessarily increasing conditioning, except in less-conditioned runners. Use of a buoyancy vest or flotation belt reduces the energy demands of DWR, but promotes better specificity by encouraging a more “cross-country” style of gait, whereby the leg sweeps back at a large angle and the foot “pushes” down at the bottom of the stride.⁶⁸ Nevertheless, a limitation of DWR is that the running motion does not always mimic running on land, and there is increased resistance to all movements due to the hydrostatic properties of water.⁶⁸ Ultimately, DWR is useful during recovery from a BSI, as it can be used to maintain conditioning without excessively loading the pathological site, but it does not replace the need to progress to dry-land running when the pathology permits.

Accelerating Tissue-Level Healing It would be beneficial to provide an athlete with a means of accelerating the healing of a BSI, particularly if it also accelerated a return to running. While there are no proven methods of accelerating BSI healing, a number of candidate methods have been proposed. Low-intensity (less than 0.1 W/cm²) pulsed ultrasound therapy currently has the greatest support, given its observed beneficial effects on complete bone fractures.¹¹⁵ Low-intensity pulsed ultrasound has been effective in stimulating union in 98% and 94% of BSIs displaying delayed union and nonunion, respectively.⁷⁹ However, the benefits of low-intensity pulsed ultrasound on acute BSIs remain unclear. Although low-intensity pulsed ultrasound has preclinical⁷⁴ and anecdotal clinical^17,62 support in the management of acute BSIs, a pilot randomized clinical trial of low-intensity pulsed ultrasound effects on recovery following tibial BSI suggested no benefit.¹⁰³ Extracorporeal shockwave therapy and electromagnetic and capacitive coupled electric fields have also been considered as modalities for promoting BSI repair; however, studies of their effects have principally been limited to recalcitrant BSIs. One randomized controlled trial did study the effects of capacitive coupled electric fields on acute tibial BSIs, but was unable to demonstrate a generalized effect on time to return to activity.⁷

A number of pharmaceutical agents may facilitate BSI healing, yet their use is unlikely to receive regulatory approval due to the difficulty of establishing their efficacy in clinical trials of BSIs. Thus, their administration in runners with a BSI will likely remain off label. Bisphosphonates that impair osteoclast-mediated bone resorption have been explored as a potential therapy for BSI. Case series have reported rapid resolution of BSI symptoms with intravenous pamidronate or ibandronate^23,110; however, this may not necessarily equate to enhanced tissue-level healing, with preclinical studies showing that bisphosphonates delay BSI healing by suppressing intracortical bone remodeling and damage removal.^66,108 In contrast, anabolic agents that more specifically target bone-forming activities may hold greater potential. Two such therapies are parathyroid hormone and antisclerostin antibody therapy. Parathyroid hormone, when administered intermittently, promotes osteoblastogenesis and osteoblast survival, whereas antisclerostin antibody therapy inhibits osteocyte secretion of sclerostin to facilitate Wnt signaling and subsequently promote osteoblast proliferation and function.²¹ Both agents stimulate bone formation and accelerate fracture healing in preclinical models. Whether the same fracture-healing benefits carry over to humans and BSI is unknown, although an initial preclinical study suggested that parathyroid hormone therapy accelerated healing of experimentally induced BSIs.¹⁰⁸

Phase 2: Return to Running

Beginning and Progressing Initial Running Graduated running programs are used in the management of low-risk BSIs to introduce controlled loading and facilitate a return to running in a timely yet safe manner. While loading is central to the development of low-risk BSIs, recovery is best met by a balance between rest from aggravating activities and performance of appropriate loading. Appropriate loading can be defined as loading that does not provoke BSI symptoms either during or after completion of an activity. Once a runner with a low-risk BSI becomes pain free during unassisted walking, the runner can start the gradual process of reintroducing running-related loads. While there is no established protocol for returning to running during recovery from a low-risk BSI, various programs have been developed based on clinical experience. One program requires the athlete to be pain free for at least a week before beginning to run every other day for 2 weeks at half the athlete's usual pace and distance.³ Thereafter, running distance and frequency are gradually increased over 3 to 6 weeks to the athlete's preinjury level. Once the athlete can run the usual training distance at the usual frequency, the pace is gradually increased. If the athlete experiences BSI symptoms at any time during incremental loading, running is ceased for at least 1 to 2 days and resumed at a lower level.

The above program provides general guidelines for resuming and progressing running activities; however, an example of a more prescriptive graduated running program to facilitate the return of a recreational runner to 30 minutes of running is provided in TABLE 3. The program consists of a pre-entry stage and 3 running stages. Once a patient is completely pain free for 5 consecutive days during usual activities, the patient may leave the pre-entry stage (stage 0) and commence deliberate progressive loading. Stage 1 introduces loading in 30-minute sessions, separated by rest days. Sessions in this stage consist of increasing durations of jogging (defined as running at 50% of normal pace) and decreasing durations of walking. The pace of running is progressed in stage 2 until the athlete can run for 30 minutes at the usual pace, with stage 3 consisting of running on 2 consecutive days followed by a rest day. The last stage incorporates individualized running until complete return to desired running activities.

Progress through each stage of the graduated running program is determined by BSI provocation. If an athlete is able to complete a session with no BSI symptoms and does not experience latent symptoms, the athlete can progress to the next level during the ensuing session. However, if BSI symptoms are experienced during a session or the athlete experiences latent symptoms, the athlete must stop the session and return to the last level that he or she was able to successfully complete at the next session. For example, if the athlete experienced symptoms while walking for 6 minutes and jogging for 4 minutes, the athlete would take a rest day, then walk for 7 minutes and jog for 3 minutes each at the next running session.

Using a graduated loading program that is strictly guided by pain to manage low-risk BSIs, the athlete and clinician can facilitate recovery while being relatively confident that there would be no progression of the pathology. Of course, this requires the full understanding of the athlete regarding the appropriate progression through the program and adherence to the set pain levels. Patients with lower-grade BSIs may be able to progress more quickly through a graduated loading program, and the total duration of their program may be reduced from that outlined. Program duration should be based on the expected time course for recovery, which in turn is dependent on the classification of their BSI. Thus, all patients should perform individualized programs.

Antigravity Treadmill Training When available, an antigravity treadmill may initially be used in the place of overground running, as ATT may allow an athlete to run at higher intensities earlier during recovery, but with lower bone loading. Thereby, ATT can be used to maintain fitness while protecting the BSI site. Antigravity treadmills consist of a treadmill with an air-filled, pressure-controlled chamber that surrounds the lower half of the body from the waist down (FIGURE 4). Pressure in the chamber is modulated to unweight the runner in 1% increments so that the runner is running with between 100% and 20% body weight. Grabowski and Kram⁵³ found that certain combinations of velocity and body-weight support provided aerobic training similar to normal-weight running, but with less peak GRFs. Based on the data, if a person normally runs at 3 m/s with no body-weight support, he or she could run at 5 m/s with 43% body-weight support and achieve equal metabolic demand but decrease peak GRF by 32%.

Download FigureDownload PowerPoint

Runners with a low-risk BSI can start ATT once pain free during walking and ADLs for at least a week. A typical starting point is to jog every other day 3 times for 5 minutes, with between 50% and 70% body weight and 1 minute of recovery between repetitions. This is performed for an initial week to acclimatize to ATT while symptoms are monitored for provocation. How to progress from this initial stage is currently somewhat of an art rather than science, as there are few published protocols guiding ATT progression. However, a proposed progression is to increase, over the next 2 weeks, the duration of each run to a total of 20 to 30 minutes, body-weight percentage by 5% to 10%, and running speed.²⁴ Thereafter, as long as there are no BSI symptoms, the progression may continue until the athlete can run at the usual speed and duration with 90% body weight. Once this is achieved, ground running is introduced and relatively quickly progressed over a couple of weeks to normal durations and intensities.

Running Gait Retraining Despite a slow, progressive return-to-running program, the persistence of faulty mechanics may hinder the healing process or contribute to the elevated risk for a repeat injury. Therefore, it is important to identify and address any underlying mechanics when treating a runner with a BSI. Based on the hypothesis that bone loading is directly related to GRF and acceleration parameters, interventions that reduce GRFs and shock during running gait may represent a means of decreasing BSI risk. By reducing the magnitude and rate of bone loading, the number of loading cycles until microdamage accumulation and fatigue failure may be increased. A number of gait retraining techniques to reduce bone loading during running are currently being investigated, including the use of biofeedback, altering stride rate, and modifying initial contact.

Use of biofeedback from an accelerometer attached to the distal tibia has been described as a potential means of encouraging reductions in loading magnitude and rate. Crowell and colleagues³¹ displayed tibial acceleration data in real time on a monitor in front of subjects running on an instrumented treadmill. Following a preliminary warm-up and baseline data-collection period, subjects were asked to “run softer” and reduce their peak positive acceleration by 50%. After 10 minutes of biofeedback followed by 10 minutes without biofeedback, 4 of 5 subjects had significantly reduced peak positive acceleration, and all subjects had reduced GRF impact peak and loading rates, demonstrating feasibility of the biofeedback technique. In a subsequent follow-up study, Crowell and Davis³⁰ provided accelerometer-based, real-time visual biofeedback to 10 runners with high peak positive tibial accelerations (greater than 8 g). After 8 training sessions over 2 weeks, there were significant reductions in peak positive acceleration and GRF loading rates. These changes, with the addition of a reduction in GRF impact peak, persisted for 1 month following feedback training completion, indicating short-term retention of the subjects' altered running form. How the subjects altered their gait to reduce tibial acceleration, the long-term permanence of the reduced acceleration, and whether the reduction in acceleration and subsequent GRF parameters alter BSI risk are areas of ongoing investigation.

Gait retraining using accelerometry biofeedback has shown promise and clinical translatability in reducing bone loading, as it does not require expensive equipment such as an instrumented treadmill. However, in the absence of accelerometry hardware, a simple alternative may be to increase a runner's stride rate above the runner's preferred rate. Increasing stride rate results in a proportional decrease in stride length (assuming a constant running speed) and positions the feet more underneath the center of mass at initial contact, with the knees and ankles in greater flexion and plantar flexion, respectively.⁵⁷ The net result is a reduction in vertical excursion and velocity of the center of mass, reduced peak hip adduction angle and moment, and reduced GRFs and tibial accelerations.^{26,33,42,56–58} Changes in stride rate can be encouraged using a metronome, with increases of 10% or less above an athlete's preferred rate being adequate to reduce impact loading while maintaining or even reducing oxygen consumption.^22,56 An increase in stride rate and concomitant decrease in stride length increase the number of loading cycles experienced when running a given distance, potentially offsetting the benefit of stride-rate changes on BSI risk. However, the reduced loading magnitude associated with decreased stride length has been modeled to outweigh any potential detriment of the increased number of loading cycles.⁴⁰

An alternative method of gait retraining to reduce impact loads has been to encourage a more FFS, rather than a heel or RFS pattern. Up to 89% of shod runners land with an RFS pattern,⁷³ possibly due to the thick, cushioned heel of modern running shoes. Landing with an RFS is associated with a defined impact peak in the vertical GRF during contact and a high rate of loading. In contrast, an FFS pattern eliminates the impact transient by reducing the effective mass at contact through eccentrically controlled dorsiflexion.⁷⁵ A potential means of promoting a more FFS pattern has been to have an athlete run barefoot or in a minimalist shoe. Barefoot running encourages a FFS pattern, as running barefoot with an RFS pattern results in very high loading rates and is typically uncomfortable.² In addition to encouraging an FFS pattern, barefoot running is associated with a shorter stride length and higher stride rate than typical shod running using an RFS pattern.² The net result is a reduction in the magnitude and rate of loading and, potentially, BSI risk. It is possible that a more FFS pattern may increase BSI risk in the metatarsals, due to greater bending forces. However, a natural barefoot landing is described typically as a very mild FFS, with the foot landing slightly plantar flexed and the heel slightly off the ground before it is actively lowered.² A low inclination of the foot at contact may reduce the bending moment on the metatarsals to reduce any potential elevation in BSI risk at this site. However, transitioning to minimal-footwear running should be done with caution. In a study by Ridge et al,¹⁰² runners transitioning to minimalist shoes with little guidance more often developed metatarsal bone marrow edema than those remaining in their traditional shoes. Therefore, transitioning should be done slowly and be associated with a strengthening program of the calf and foot intrinsic muscles.

The benefit of the aforementioned gait retraining techniques remains an area of active inquiry, and the induction of running gait changes should not be taken lightly. Runners have typically been running with their particular gait pattern for many years, and the ability of a runner's musculoskeletal system to adapt to the nuances of their particular gait should not be underestimated. However, a history of repeat BSI and the accumulating loss of running time are signs that gait retraining should be considered. In doing so, it must be remembered that when inducing a change in gait, there is always the potential of altering injury risk at an alternative site. Thus, transitioning to a new running gait should be performed slowly and may be benefited by a preconditioning program.

Running Program Design Incrementing a running program too rapidly or frequently relative to a runner's usual activities can contribute to the generation of a BSI by upsetting the balance between bone microdamage formation and its removal. Unfortunately, there is no accepted algorithm for how much an individual can modify a running program before excessively heightening the risk of BSI. The historical rule of thumb has been to change a running program by no more than 10% per week to reduce injury risk; however, this “rule” is not generalizable on an individual basis, as different runners are able to tolerate more or less change before developing an injury.¹⁹ In order to provide a runner returning from a BSI with appropriate advice with regard to the design and advancement of a running program, it is important that the runner recall as much as possible about the preinjury training regime. This should include information on changes in any feature that may have altered the load being introduced to the skeleton, including training intensity, duration, frequency, type, surface, technique, shoes, and so on. In addition, recovering runners should be encouraged to maintain a training diary containing these data to not only track running progress but also to provide reliable data regarding the running program for future reference. There are numerous useful apps for mobile devices that can semi-automatically collect some of these data. In addition to limiting the magnitude and number of running variables changed at any given time, the use of cyclic training methods should be encouraged. Cyclic training may involve the introduction of rest periods into a training program, or the replacement of overground running sessions with lower bone-loading running activities, such as DWR and ATT. This may involve a monthly regime of 3 weeks of running and 1 week of no or low-load activities.