Secondary Logo

Journal Logo

Treatment and Prognosis of High- and Low-Risk Kaeding Grade II Bone Stress Injuries

Curell, Angela M.1; Nye, Nathaniel S.2; Webber, Bryant J.2; Pawlak, Mary T.2; Boden, Barry P.3

Translational Journal of the American College of Sports Medicine: August 1, 2019 - Volume 4 - Issue 15 - p 114–118
doi: 10.1249/TJX.0000000000000089
Original Investigation

ABSTRACT Bone stress injuries (BSI) may be classified as high- or low-risk based on the anatomic location of injury and by grade based on severity. Kaeding grade II (K-GII) BSI are characterized by symptomatic marrow or periosteal edema without a fracture line. This retrospective cohort study aims to compare outcomes between high- and low-risk K-GII BSI. We hypothesize that patients with high- and low-risk K-GII BSI experience similar recovery rates. Data were collected via chart review on all patients at a primary care clinic with a magnetic resonance imaging–confirmed diagnosis of K-GII BSI during a 15-month surveillance period. High- and low-risk patients were compared for two primary outcomes—time to become asymptomatic and time to return to activity—and for the secondary outcome of treatments received. A total of 129 K-GII BSI were sustained by 87 patients. For all patients diagnosed with a K-GII BSI, the mean time to become asymptomatic was 40 ± 27 d, and the mean time to return to activity was 49 ± 31 d. There was no difference in time to become asymptomatic (P = 0.762) or to return to activity (P = 0.164) between high-risk (n = 20) and low-risk (n = 67) patients. Treatment modalities were similar between the two groups. All K-GII BSI healed with nonoperative treatment at similar rates, regardless of classification as high- or low-risk, to include tarsal navicular and talus BSI. Early identification of BSI, before progression to a fracture line, leads to favorable results for both high- and low-risk K-GII BSI.

1San Antonio Uniformed Services Health Education Consortium, Ft. Sam, Houston, TX

2Trainee Health Squadron, Joint Base San Antonio–Lackland, TX

3The Orthopaedic Center, Division of Centers for Advanced Orthopaedics, Rockville, MD

Address for correspondence: Angela M. Curell, 3551 Roger Brooke Drive, Fort Sam, Houston, TX 78234 (E-mail:

Back to Top | Article Outline


Bone stress injuries (BSI) produce significant morbidity in populations that engage in repetitive and intense weight-bearing exercise, such as competitive athletes and military trainees (1–4). Suspected BSI are often evaluated with imaging, which may enhance diagnosis and treatment. Several imaging-based grading classifications have been proposed, such as the Fredericson system for tibial stress fractures (5), but most are location specific or fail to incorporate clinical factors. Kaeding and colleagues recently validated a simple, standardized classification scheme for all types of BSI, based on both radiographic and clinical findings. In the Kaeding system, grade I is an asymptomatic stress reaction, grade II indicates pain and corresponding marrow or periosteal edema but no fracture line, grade III is a nondisplaced fracture, grade IV denotes displaced fracture, and grade V indicates a nonunion (see image, Supplemental Digital Content, MRI demonstrating Kaeding grade II BSI, (6).

BSI may be further categorized as either low-risk or high-risk based on the prognostic implications associated with the anatomic site of injury (7,8). Low-risk BSI usually heals completely without surgical intervention; common examples include the femoral shaft, tibia, fibula, calcaneus, and the first through fourth metatarsals. BSI of the patella, anterior cortex of the tibia, tension side of the femoral neck, medial malleolus, talus, tarsal navicular, fifth metatarsal, and great toe sesamoids are classified as high-risk (3), as they may progress to complete fracture or nonunion without strict adherence to weight-bearing restrictions and often require surgical intervention (7,9–13).

Although substantial research has been conducted on BSI prevention, epidemiology, and treatment (2,3,10,14–20), most studies classify all BSI as a single disease entity, irrespective of grade or risk level. Despite being common injuries, Kaeding grade II (K-GII) BSI data are lacking. This retrospective cohort study evaluates outcomes and treatments of K-GII injuries in a large military training environment.

Back to Top | Article Outline


Cases of K-GII BSI among patients seen at a primary care clinic at Joint Base San Antonio–Lackland, Texas, were identified through the Trainee Health Surveillance stress fracture registry, which maintains demographic and injury data for BSI diagnosed among trainees on the installation. The database was queried for all K-GII BSI diagnosed between October 1, 2014, and December 31, 2015. A local BSI treatment policy, which prioritized magnetic resonance imaging (MRI) as the advanced imaging modality of choice, was implemented in September 2014. Skeletal scintigraphy had been used almost exclusively before this policy due to fiscal and logistical considerations (21).

For all incident BSI during the surveillance period, chart reviews were conducted in the Armed Forces Health Longitudinal Technology Application, the outpatient electronic health record of the U.S. Military Health System. Using clinical notes and radiologist interpretations of MRI findings, all BSI were classified according to Kaeding’s system to verify appropriate grading (6). For all K-GII BSI, the following variables were obtained: age; sex; disposition (i.e., return to activity or medical discharge); high-risk or low-risk bone involvement (7,8); symptom onset and resolution dates; dates of removal from and return to activity or discharge; running restriction start and stop dates; walk-to-run program start and stop dates, if applicable; and additional treatments rendered. Treatments included calcium/vitamin D supplementation (1200 mg calcium carbonate and 800 IU vitamin D3 by mouth daily, or cholecalciferol 50,000 IU orally once weekly, or both), analgesics (nonsteroidal anti-inflammatory drugs or acetaminophen), crutches, brace, and controlled ankle motion (CAM) boot. Some patients were also enrolled in a walk-to-run program with gait training. To graduate from the program, a patient had to run continuously at ≥6 mph, without pain, for 30 min.

Symptom onset was defined as the date the patient reported first experiencing pain at the injury site. Symptom resolution was defined as the date on which the patient could perform activities of daily living without pain and then remained asymptomatic with activity thereafter. Treatment start and stop dates were determined from clinical notes in the electronic health record. Treatment stop dates were determined either by clear documentation of the decision to discontinue treatment or, when not explicitly stated, by the date of the first note that did not document use of the treatment. Because the tibial shaft and tibial plateau were the most common BSI sites, a post hoc analysis was performed to compare demographics, symptoms, treatments, and outcomes of these two groups. Patients with BSI of both locations were excluded.

Descriptive statistics, including mean ± SD for continuous variables, were used to characterize our population. Student’s independent t-test, one-way ANOVA, and chi-square test were used to compare high-risk and low-risk BSI according to the following variables: age; sex; disposition; walk-to-run disposition (graduated, failed, and did not start); time from symptom onset to resolution; time to return to activity; time in a “no-run” status; time to complete the walk-to-run program; and presence or absence of treatment modalities. Patients who sustained at least one high-risk BSI, with or without one or more concomitant low-risk injuries, were classified as “high-risk,” whereas those who sustained one or more low-risk BSI but no high-risk injury were classified as “low-risk.”

Statistical analyses were performed using OpenEpi v3.01; after applying the Bonferroni correction for multiple comparisons, two-tailed P values <0.004 were considered statistically significant.

This study was approved by the 59th Medical Wing Institutional Review Board.

Back to Top | Article Outline


A total of 129 K-GII BSI were sustained by 87 patients during the 15-month surveillance period. The population was predominantly male (66%) with a mean ± SD age of 21 ± 2.8 yr (Table 1). Ninety-seven (75%) BSI were sustained in low-risk locations, and 67 (77%) patients were classified as low-risk because they had no concomitant high-risk BSI. The most common injury locations were the tibial shaft (n = 32) and proximal tibial metaphysis (n = 21) (Table 2). Thirty-two BSI (25%) were sustained in high-risk locations, with the predominant injury locations being the talus (n = 15) and navicular (n = 12). Twenty patients (23%) had at least one BSI in a high-risk bone. Patients classified as low-risk had a mean of 1.5 low-risk BSI, whereas patients classified as high-risk had a mean of 1.6 high-risk BSI and 2.6 low-risk BSI (Table 2). Forty-seven patients had one BSI, 18 patients had two BSI, 12 patients had three BSI, 3 patients had four BSI, and the remaining 7 patients had greater than four BSI. Among the 30 patients with multiple BSI, the most frequent combinations were tibial plateau and femoral condyle (n = 8), tibial plateau and tibial shaft (n = 7), tibial plateau and distal fibula (n = 3), and cuneiforms and first through fourth metatarsals (n = 3). Among those with three BSI, 8 (67%) had tibial involvement and 3 (25%) had cuneiform involvement. Males (n = 57) and females (n = 30) had similar breakdowns of BSI locations; 10 females (33%) and 10 males (20%) were classified as having high-risk BSI (P = 0.167).





Patients required a mean ± SD of 40 ± 27 d to become asymptomatic and 49 ± 31 d to return to activity, with no significant difference between high-risk and low-risk (P = 0.762 and 0.164, respectively). This statistical equivalence persisted after stratifying high-risk BSI by anatomic site (Table 3). Patients were on a running restriction profile from the time of diagnosis for 35 ± 20 d and required 22 ± 19 d to complete a walk-to-run program, also with no significant difference between high- and low-risk groups (Table 1).



Of the 56 patients who began a walk-to-run program, 46 (82%) completed the program; return to activity or discharge after starting a walk-to-run program was not associated with risk level (P = 0.865). Most patients who failed to start the program were discharged before they had healed sufficiently to begin the walk-to-run program. A wide variety of treatments were used, including calcium and vitamin D supplementation (87%), crutches (72%), and analgesic medication (53%). Patients with high-risk BSI were more likely to receive a CAM boot (40%) than those with low-risk BSI (4%) (P < 0.001). Treatments rendered were otherwise equivalent between the groups (Table 1), and no patients required surgery.

Fifty patients (57%) completed rehabilitation locally and returned to activity, with no significant difference between those with high-risk (60%) and low-risk (57%) injuries (P = 0.79). The remaining patients (n = 37) were discharged for a variety of reasons, including low motivation to return to activity, administrative issues, medical issues unrelated to stress injury, and failure to progress with rehabilitation. None demonstrated a fracture line at the time of discharge or required surgery before discharge. All patients with a high-risk BSI were cleared for military discharge (i.e., were determined safe to travel and did not require urgent surgery) by the orthopedic clinic. The average time from diagnosis to discharge for high-risk patients was 36.1 d. At the time of discharge, 62.5% of high-risk patients had started a walk-to-run program, 2 were using a CAM boot, and 1 was using crutches. A greater percentage of males (47%) than females (33%) were discharged, but the difference was not statistically significant (P = 0.209). Similar proportions of males and females were discharged (P = 0.209) and completed the walk-to-run program (P = 0.312), and they had similar symptom duration (P = 0.483), time out of training (P = 0.332), and time on no-run status (P = 0.305).

After excluding those with BSI of both the tibial shaft and the plateau, patients with BSI of either location were similar by age (P = 0.702), sex (P = 0.945), symptom duration (P = 0.560), and treatments received. However, 10 (71%) of 14 patients in the tibial plateau group were medically discharged, compared with 4 (17%) of 24 patients in the tibial shaft group (P < 0.001).

Back to Top | Article Outline


In a sample of patients who sustained K-GII BSI, rates of healing and return to activity were equivalent regardless of injury location and risk level. Furthermore, subjects in this study required much shorter healing and rehabilitation times than those seen in previous BSI studies. The mean return to activity time of 49 d (43 d for high-risk and 52 d for low-risk) was much shorter than times reported elsewhere. Matheson and colleagues (22) observed that athletes with conservatively managed BSI (N = 320) had a mean recovery time of 90 d. Swenson and colleagues (23) found that athletes with tibial stress fractures treated without a brace (n = 9) required a mean of 77 d to return to play. Dobrindt and colleagues (24) reported a mean recovery time of 95 d for low-grade lesions without fracture line (n = 52). The shorter recovery time in our study may reflect earlier detection of the BSI, differences in the population, such as injury severity and healing ability, or differences in the environment. Notably, BSI studies of civilian athletes are typically conducted at the collegiate and professional levels, where the physical demands of returning to play are very high. By contrast, returning to activity in our population required a patient to be pain-free and have completed a walk-to-run program.

BSI of some anatomic sites are labeled “high-risk” due to greater probability of progression to complete fracture, delayed union, and nonunion (7). Tarsal navicular stress fractures, in particular, are often missed on routine radiographs because of the oblique positioning of the bone, delaying the diagnosis and increasing the likelihood of nonunion and need for surgical management (25). This study confirms previous evidence that early diagnosis of tarsal navicular BSI before fracture line development carries an excellent prognosis and may be treated conservatively with activity restriction and limited weight bearing for a few weeks (26). Fifth metatarsal and medial malleolus BSI also have high rates of delayed union and nonunion when diagnosed belatedly (25,27). In this cohort of patients diagnosed at the K-GII stage, fifth metatarsal and medial malleolus BSI had an excellent prognosis. However, further research is necessary to confirm these results due to the low numbers in this study.

This study confirms earlier findings that talar BSI are usually associated with at least one other BSI (28). The classification of talus BSI as high- or low-risk has been controversial (7,29–31). This may be due to the rarity of talar stress fractures, the multitude of locations within the talus where these BSI can occur, and the lack of staging these BSI in the literature. In this report, we classified talus stress fracture as high-risk due to the possibility of poor outcomes with this BSI. However, the outcomes demonstrated that BSI of the talus resolved at a rate similar to all low-risk BSI, indicating that talar BSI diagnosed early in their clinical course have a favorable prognosis.

This study also provides BSI data regarding specific tibial locations. Although statistical power is limited because of small sample sizes, it seems that outcome data of patients with BSI of the tibial shaft and tibial plateau are largely similar. However, it is unclear why those with a tibial plateau BSI in this cohort were more likely to be medically discharged from training.

Although the findings are limited by incomplete follow-up, which reduces the statistical power to identify potential differences between the groups for some outcomes, the similar discharge percentage between those with high- and low-risk BSI suggests that misclassification is nondifferential and thus would likely bias results toward the null. In addition to larger-scale studies with greater statistical power, further research should be conducted on K-GII BSI that did not occur in this population, including tension side of the femoral neck, patella, anterior cortex of the tibia, and great toe sesamoids.

Overall, this study demonstrates that high- and low-risk BSI have comparably favorable prognoses when detected early, indicating that the grade at diagnosis may be more important than the location. This finding contradicts literature citing prognostic differences between high- and low-risk stress injuries (7,8,24,32). These previous studies, however, were largely conducted before the introduction of the Kaeding classification system and contained significant numbers of advanced BSI with an identifiable fracture line. Our study eliminated this selection bias by restricting to K-GII cases. The discrepancies between our findings and those in previous studies indicate a more favorable prognosis for K-GII injuries compared with more advanced injuries, suggesting that BSI grading is critical in management. It also supports common clinical reasoning that earlier detection of both high- and low-risk BSI will reduce recovery time and the need for surgical intervention.

No financial disclosures were reported by the authors of this article. This project was completed without grants; however, Department of Defense equipment and facilities were used in data collection, analysis, and writing.

The views expressed are those of the authors and do not reflect the official views or policy of the Department of Defense and its components. The results of the present study do not constitute endorsement by the American College of Sports Medicine. The authors have no conflicts of interest to declare.

Back to Top | Article Outline


1. Rizzone KH, Ackerman KE, Roos KG, et al. The epidemiology of stress fractures in collegiate student-athletes, 2004–2005 through 2013–2014 academic years. J Athl Train. 2017;52(10):966–75.
2. Kaufman KR, Brodine S, Shaffer R, et al. Military training-related injuries: surveillance, research, and prevention. Am J Prev Med. 2000;18(Suppl 3):54–63.
3. Knapik J, Ang P, Reynolds K, et al. Physical fitness, age, and injury incidence in infantry soldiers. J Occup Med. 1993;35(6):598–603.
4. Lee DArmed Forces Health Surveillance Center (AFHSC). Stress fractures, active component, U.S. Armed Forces, 2004–2010. MSMR. 2011;18(5):8–11.
5. 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. 1995;23(4):472–81.
6. Kaeding CC, Miller T. The comprehensive description of stress fractures: a new classification system. J Bone Joint Surg Am. 2013;95(13):1214–20.
7. Boden BP, Osbahr DC. High-risk stress fractures: evaluation and treatment. J Am Acad Orthop Surg. 2000;8(6):344–53.
8. Boden BP, Osbahr DC, Jimenez C. Low-risk stress fractures. Am J Sports. 2001;29(1):100–11.
9. Cowan DN, Bedno SA, Urban N, et al. Step test performance and risk of stress fractures among female army trainees. Am J Prev Med. 2012;42(6):620–4.
10. Dickson TB Jr, Kichline PD. Functional management of stress fractures in female athletes using a pneumatic leg brace. Am J Sports Med. 1987;15(1):86–9.
11. Ivkovic A, Bojanic I, Pecina M. Stress fractures of the femoral shaft in athletes: a new treatment algorithm. Br J Sports Med. 2006;40(6):518–20.
12. Jacobsson J, Timpka T, Kowalski J, Nilsson S, Ekberg J, Renström P. Prevalence of musculoskeletal injuries in Swedish elite track and field athletes. Am J Sports Med. 2012;40(1):163–9.
13. Sharma J, Greeves JP, Byers M, Bennett AN, Spears IR. Musculoskeletal injuries in British Army recruits: a prospective study of diagnosis-specific incidence and rehabilitation times. BMC Musculoskelet Disord. 2015;16:106–13.
14. Cosman F, Ruffing J, Zion M, et al. Determinants of stress fracture risk in United States Military Academy cadets. Bone. 2013;55(2):359–66.
15. Dash N, Kushwaha A. Stress fractures-a prospective study amongst recruits. Med J Armed Forces India. 2012;68(2):118–22.
16. Jacobs JM, Cameron KL, Bojescul JA. Lower extremity stress fractures in the military. Clin J Sports Med. 2014;33(4):591–613.
17. Jones BH, Thacker SB, Gilchrist J, et al. Prevention of lower extremity stress fractures in athletes and soldiers: a systematic review. Epidemiol Rev. 2002;24(2):228–47.
18. Kupferer KR, Bush DM, Cornell JE, et al. Femoral neck stress fracture in Air Force basic trainees. Mil Med. 2014;179(1):56–61.
19. Nye NS, Pawlak MT, Webber BJ, Tchandja JN, Milner MR. Description and rate of musculoskeletal injuries in air force basic military trainees, 2012–2014. J Athl Train. 2016;51(11):858–65.
20. Valimaki VV, Alfthan H, Lehmuskallio E, et al. Risk factors for clinical stress fractures in male military recruits: a prospective cohort study. Bone. 2005;37(2):267–73.
21. Nye NS, Covey CJ, Sheldon L, et al. Improving diagnostic accuracy and efficiency of suspected bone stress injuries. Sports Health. 2016;8(3):278–83.
22. 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. 1987;15(1):46–58.
23. Swenson EJ Jr, DeHaven KE, Sebastianelli 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(3):322–8.
24. Dobrindt O, Hoffmeyer B, Ruf J, et al. Estimation of return-to-sports-time for athletes with stress fracture—an approach combining risk level of fracture site with severity based on imaging. BMC Musculoskelet Disord. 2012;13:139–46.
25. 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(5):700–12.
26. Torg JS, Moyer J, Gaughan JP, Boden BP. Management of tarsal navicular stress fractures: conservative versus surgical treatment: a meta-analysis. Am J Sports Med. 2010;38(5):1048–53.
27. Orava S, Karpakka J, Taimela S, Hulkko A, Permi J, Kujala U. Stress fracture of the medial malleolus. J Bone Joint Surg Am. 1995;77(3):362–5.
28. Niva MH, Sormaala MJ, Kiuru MJ, Haataja R, Ahovuo JA, Pihlajamaki HK. Bone stress injuries of the ankle and foot: an 86-month magnetic resonance imaging-based study of physically active young adults. Am J Sports Med. 2007;35(4):643–9.
29. Bradshaw C, Khan K, Brukner P. Stress fracture of the body of the talus in athletes demonstrated with computer tomography. Clin Sports Med. 1996;6(1):48–51.
30. McInnis KC, Ramey LN. High-risk stress fractures: diagnosis and management. PM&R. 2016;8(Suppl 3):S113–24.
31. Sormaala MJ, Niva MH, Kiuru MJ, Mattila VM, Pihlajamäki HK. Outcomes of stress fractures of the talus. Am J Sports Med. 2006;34(11):1809–14.
32. Kahanov L, Eberman LE, Games KE, et al. Diagnosis, treatment, and rehabilitation of stress fractures in the lower extremity in runners. Open Access J Sports Med. 2015;6:87–95.

Supplemental Digital Content

Back to Top | Article Outline
© 2019 American College of Sports Medicine