Running-Related Injuries Captured Using Wearable Technology during a Cross-Country Season: A Preliminary Study : Translational Journal of the American College of Sports Medicine

Secondary Logo

Journal Logo

Observational Trial

Running-Related Injuries Captured Using Wearable Technology during a Cross-Country Season: A Preliminary Study

DeJong Lempke, Alexandra F.1; Hart, Joseph M.2; Hryvniak, David J.3; Rodu, Jordan S.4; Hertel, Jay5

Author Information
Translational Journal of the ACSM 8(1):e000217, Winter 2023. | DOI: 10.1249/TJX.0000000000000217



Lower extremity running-related injuries are prevalent among collegiate competitive runners, with injury rates reported in between 3.96 and 5.85 per 1000 athlete exposures (1,2). Repetitive stress injuries constitute approximately half of all collegiate running-related injuries, which are often attributed to aberrant biomechanical profiles (1,2). Laboratory-based assessments have identified that decreased cadence, increased stride length and loading, and increased lower extremity frontal plane motion are associated with some of the most prevalent repetitive stress injuries (3–6). Although these controlled assessments provide a framework for understanding the mechanism of running-related injuries, prospective studies assessing running biomechanics among athletes who develop injuries in typical training settings are needed (7,8).

Wearable sensors allow for robust outdoor gait assessments to collect significantly more data than has historically been possible using traditional laboratory study designs (9,10). Validated sensors specific to measuring running biomechanics have been used to assess runners currently experiencing lower extremity injury symptoms and have unearthed distinct biomechanical adaptations that have not been identified through treadmill-based assessments (11,12). Researchers have also used wearable technology to identify loading-based risk factors among male collegiate runners who developed injuries over the course of a competitive track season (13). These studies suggest that wearables may be used to identify deviations in runners’ movement patterns leading up to injury incidence through prospective assessments.

We previously conducted a study among collegiate runners who were healthy at the initiation of a competitive cross-country season to assess sensor-derived biomechanical patterns as they related to session ratings of perceived exertion (sRPE) and general wellness measures (i.e., sleep quality, stress, mood) (10). There was a subset of athletes in this sample who developed lower extremity, repetitive stress injuries over the course of the study timeframe. Given that we collected sensor data on these athletes leading up to their injuries, we have the opportunity to evaluate biomechanical risk factors as they relate to specific injury cases. Therefore, the purpose of this preliminary, observational study was to present in situ biomechanical profiles among collegiate cross-country runners who developed running-related injuries compared to healthy team measures, with a particular emphasis on similar injury cases incurred throughout the season.


This assessment was conducted as a part of a larger, prospective study among Division 1 collegiate male and female cross-country athletes (10). To be included in the study, participants had to be actively training in varsity cross-country practices and free from any lower extremity musculoskeletal injuries within 3 months of the start of the study. All participants provided informed consent before study procedures. The study was approved by the University of Virginia Institutional Review Board (IRB No. 21756).

Data Collection

Biomechanical data were collected over the course of the season using RunScribe Plus™ wearable sensors (Scribe Labs, Inc., Half Moon Bay, CA), which have been previously validated against gold-standard laboratory motion capture equipment (14–16). Specifically, the sensors have demonstrated concurrent validity for spatiotemporal (intraclass correlation coefficients (ICC), 0.86–0.93), kinetic (ICC, 0.89–0.92), and kinematic measures (ICC, 0.57–0.74) that were included in the present study (14–16). The sensors were lace-mounted on each runner’s left and right shoes and calibrated through a 400-m run around a standard track at the start of the season. Sensor data were recorded twice per week during sustained runs in which the entire team practiced together. This collection schema resulted in one recorded long run and one recovery run per week over the course of the 12-wk competitive season (August–November 2019). Wellness data were collected through a Qualtrics™ (Qualtrics XM, Provo, UT) survey at the conclusion of each recorded training run (10). Survey questions were based on 5-point wellness surveys previously implemented in team sport settings, in which higher scores indicate feeling subjectively better (17,18). Survey components included questions about weekly mileage, sleep quantity and quality, stress level, mood, soreness, and Borg’s 10-point rating of perceived exertion (17–20).

Identification of Cases

Athletes who indicated on their postrun surveys that they were either “experiencing an increase in tightness and soreness” or “very sore” were presented with another set of questions asking which body parts were sore and if their soreness was progressing to pain. If the athletes indicated their soreness was developing to pain, the primary investigator retrieved injury information from one of the on-site athletic trainers. We defined an injury as any case that was currently being evaluated or treated by the team athletic trainers, including time-loss and non–time-loss cases. Athletes who sustained injuries during the competitive season and continued to compete (i.e., non–season-ending injuries) were asked to fill out the Wisconsin Running Injury and Recovery Index immediately after the wellness survey to monitor recovery and return to running (21).

Data Processing

Step-by-step spatiotemporal (pace, cadence, contact time, stride length, foot strike type), kinematic (pronation excursion and maximum velocity), and kinetic (impact g, braking g, shock) data were extracted from the associated sensor’s online dashboard account for all athletes’ recorded runs. Operational definitions of these validated sensor-derived outcomes have been published elsewhere (11). Data were extracted using the software version 3.0.0 in 2020. Walking and standing events were visually identified in the data sets from when an athlete’s flight ratio fell to zero, and were removed from analyses (12,22).

The cross-country season was sectioned into preseason (weeks 1–3), early season (weeks 4–6), late season (weeks 7–9), and championship (weeks 10–11) time points to account for training fluctuations. Means and SD of sensor-derived biomechanics were calculated for healthy male and female runners and for individual injured runners for all runs recorded within each quarter of the season. These measures were used to calculate z scores of each biomechanical outcome to assess male and female team and injured runner case variability:

zscores=mean session scoremean score across seasonSDof session score

Survey responses were extracted from the Qualtrics dashboard, and string responses were numerically coded for analyses. The ratings of perceived exertion were multiplied by the session duration to calculate sRPE for each recorded run (20). Composite wellness z scores were calculated by combining the sleep quality, mood, stress, and soreness responses for each athlete.

Statistical Analyses

Descriptive analyses were used to compare weekly mileage and sRPE across the season between healthy and injured runner cases. Biomechanical and wellness z scores were assessed for injured runner cases and compared against healthy male and female group z scores with 95% confidence intervals. We specifically examined the several recorded days leading up to injury to determine if there were notable patterns outside of healthy team ranges that may have contributed to injury development.


Eight injuries were reported throughout the season (four male, four female), with four categorized as bone stress injuries (two sacral stress fractures, one male, one female; one femoral neck stress fracture, female; one fifth metatarsal stress fracture, female) and four as soft tissue injuries (two hamstring strains, one male, one female; one medial tibial stress syndrome, male; one plantar fasciitis, male). Demographics, survey responses, and injury details can be found in Table 1. Representative bone stress and soft tissue injury case weekly mileage, sRPE, wellness z scores, and biomechanics z scores compared with healthy team averages are presented in Figs. 1 and 2, and all remaining injury cases can be found in Supplemental Content 1 (figures:,,,,,

TABLE 1 - Injury Cases Recorded Throughout the Competitive Cross-Country Season.
Wisconsin RRI After Injury (Recorded Days After Injury)
Injury Type Sex Injury Date RTS? 1 2 3 4 5 6 7 8 9 10
Sacral stress fracture F 9.25.19 N
Femoral neck stress fracture F 9.18.19 N
Sacral stress fracture M 10.02.19 N
Metatarsal stress fracture F 11.04.19 N
Hamstring strain M 9.18.19 N
Hamstring strain F 8.31.19 Y 41.7% 55.6% 72.2% 63.9% 69.4% 72.2% 83.3% 72.2% 86.1% 91.7%
Medial tibial stress syndrome M 10.09.19 Y 52.8% 55.6% 55.6% 58.3% 69.4%
Plantar fasciitis M 9.15.19 Y 69.4% 83.3% 91.7% 100% 100% 100% 100% 100% 100% 100%
Injury survey data are presented solely for the athletes who returned to running training and competition during the season. The percentage indicates to what extent the athletes reported being recovered throughout their postinjury timeline in the days after injury.
F, female; M, male; N, no; RRI, running-related injury; RTS, return to sport; Y, yes.

Figure 1:
Female sacral stress fracture injury case compared with female runners who remained healthy during the cross-country season. Female sacral stress fracture case gait biomechanics z scores, wellness, and weekly mileage plotted against healthy female teammates across the season up to the point of injury (gray shaded box with dotted outline). This injury was season ending, and thus, metrics are not present for the injured runner after the point of injury. Metrics outside of the team’s 95% confidence intervals (CI) within two recorded dates preceding injury are denoted with asterisks. Long runs (LR) and recovery runs (RR) are indicated next to the respective dates on the x axis.
Figure 2:
Male hamstring strain injury case compared with male runners who remained healthy during the cross-country season. Male hamstring strain case gait biomechanics z scores, wellness, and weekly mileage plotted against healthy male teammates across the season up to the point of injury (gray shaded box with dotted outline). This injury was season ending, and thus, metrics are not present for the injured runner after the point of injury. Metrics outside of the team’s 95% confidence intervals (CI) within two recorded dates preceding injury are denoted with asterisks. Long runs (LR) and recovery runs (RR) are indicated next to the respective dates on the x axis.

Bone Injuries

Both female sacral and femoral stress fracture cases presented with increased contact time, pronation excursion and maximum velocity, impact g, and shock, but decreased cadence and stride length compared with the healthy female team ranges within two recorded days leading up to injury (Fig. 1; Supplemental Content 1a, figure, Similarly, the male sacral stress fracture case presented with increased impact g, shock, and pronation excursion and maximum velocity, with decreased cadence within two recorded runs preceding injury (Supplemental Content 1b, figure, The final bony injury was an isolated fifth metatarsal stress fracture that presented with increased cadence, contact time, pronation excursion and velocity, and impact g, but lower braking g leading up to injury (Supplemental Content 1c, figure,

Weekly mileage was lower on average within the 2 weeks leading up to bone injury cases, and several of the injured runners reported higher wellness z scores indicative of feeling subjectively better within the several days leading up to injury (Fig. 1; Supplemental Content 1b, figure, The sRPE outcomes for injured and healthy cases were largely comparable. There were no bone injury cases in which the athletes returned to participation in the same season.

Soft Tissue Injuries

Both hamstring strain cases presented with increased stride length in the recorded day preceding injury compared with healthy male and female team ranges (Fig. 2; Supplemental Content 1d, figure, Although the medial tibial stress syndrome and plantar fasciitis injuries affected different anatomical structures, both of these male lower limb, soft tissue injury cases presented with increased pronation maximum velocity and decreased shock and braking g within 2 recorded days of injury compared with the healthy male team range (Supplemental Contents 1e and 1f, figures, and

Only the female hamstring injury case reported higher weekly mileage compared with healthy female athletes (Supplemental Content 1d, figure,; however, mileage was generally consistent for the remaining soft tissue injury cases (Fig. 2; Supplemental Contents 1e and 1f, figures, and Wellness scores and sRPE scores were generally higher for soft tissue injury cases in the several days leading up to injury. Only the male hamstring strain case resulted in time lost for the remainder of the season. The remaining athletes with soft tissue injuries reported 69%–100% recovery after injury (Table 1).


Through this assessment, we identified biomechanical deviations in the days leading up to injury among select repetitive stress lower extremity injury cases that extended beyond healthy team normative measures. There were several commonalities among injuries including deviations in contact time, cadence, stride length, loading, and pronation maximum velocity from the healthy team ranges. We found that these alterations were dependent on the injury type and anatomical location. Although our study represents an initial approach to describing biomechanical factors associated with injury among cross-country athletes, these findings demonstrate the merit of incorporating athlete monitoring measures to improve clinical assessments and interventions for repetitive stress injuries.

Although the recorded injuries occurred at varying time points and were specific to the individual athletes, there were some similarities across cases that provide a lens into potential measures of interest for injury risk. The most frequently noted differences in the days leading up to bone stress injuries were increased contact time, loading, and pronation measures along with decreased cadence and stride length. Soft tissue injuries most frequently had increased stride length and pronation maximum velocity with decreased loading metrics. These biomechanical alterations were largely aligned with previous laboratory-based findings among runners with repetitive stress injuries demonstrating increased loading (23–25), larger kinematic deviations at the foot and ankle complex (6,26,27), and altered spatiotemporal patterns (28–31). As such, these parameters may represent targets for future assessments and interventions to reduce the risk of running-related injury.

Beyond laboratory-based assessments, our in situ biomechanical findings coincide with recent work incorporating wearable sensors into outdoor running monitoring in runners with existing injuries and among athletes with injury risk factors (11,12,32). In conjunction with this past work, our findings suggest that biomechanical changes in the days leading up to the reported injury may contribute to injury development and not only occur as a result of active symptoms. Although this case series offers a preliminary foundation of the hypothesis that biomechanical assessments may be useful in signaling injury cases, more robust study designs with larger patient populations are needed to substantiate these associations. With larger databases of runners, we envision that the future of running medicine could include the ability to identify individual runners who exhibit changes in metrics of injury risk that are outside the expected range of biomechanical and wellness measures (33). Through this individualized approach to monitoring, athletic trainers, physical therapists, and other key stakeholders can target intervention or injury prevention strategies to move toward individualized medicine.

Surprisingly, wellness scores were somewhat higher across injuries, indicative of feeling subjectively better before injury, which suggests that clinicians may not be necessarily able to flag athletes that are at risk for injury through subjective questionnaires alone. This finding aligns with recent work suggesting the limited predictive capability of Likert-based wellness assessments and with our previous assessment suggesting that biomechanical measures are not strongly associated with wellness measures (10,34). Furthermore, athletes largely reported decreased mileage leading up to injury compared with team averages. It is possible that athletes adjusted their training volume because of soreness or pain leading to injury, although this association is speculative.


There were several limitations to this study. This was an observational assessment of individual injuries, and as such should not be extrapolated to the greater population. Instead, this study sets the foundation for future prospective research. We were unable to assess other underlying physiological factors that contribute to the overall athlete injury model, including factors such as heart rate, nutrition, and other biomarkers. We were only able to record runs twice per week; future work should assess different training regimens to evaluate factors associated with injury during these activities and continue to investigate the significance of cumulative loading among cross-country athletes. We acknowledge that other biomechanical measures may contribute to running-related injury; however, we were limited to assessing the specific sensor-derived outcomes. Calculation of the variables derived from the sensors is conducted onboard the devices and is proprietary to the software, and as such, we are unable to report these underlying processes.


Several gait biomechanics measures, including stride length, loading, cadence, contact time, and pronation maximum velocity, were found to differ among injured athletes in the days leading up to injury in collegiate cross-country runners as compared with healthy team measures. This study offers a foundation for prospective assessments in competitive runners to assess injury risk.

We would like to thank the athletes for their participation in this research study, and the coach and athletic training staff for their support. The results of this study do not constitute endorsement by the American College of Sports Medicine.

Dr. DeJong Lempke is an Associated Personnel with Boston Children’s Hospital and has pending grant funding from VALD Performance for a separate project. There are no other pertinent conflicts of interest or funding sources to declare.


1. Kerr ZY, Kroshus E, Grant J, et al. Epidemiology of National Collegiate Athletic Association men’s and women’s cross-country injuries, 2009–2010 through 2013–2014. J Athl Train. 2016;51(1):57–64.
2. Chandran A, Morris SN, Boltz AJ, et al. Epidemiology of injuries in National Collegiate Athletic Association women’s cross-country: 2014–2015 through 2018–2019. J Athl Train. 2021;56(7):622–8.
3. Futrell EE, Jamison ST, Tenforde AS, Davis IS. Relationships between habitual cadence, footstrike, and vertical load rates in runners. Med Sci Sports Exerc. 2018;50(9):1837–41.
4. Ferber R, Davis IM, Hamill J, et al. Kinetic variables in subjects with previous lower extremity stress fractures. Med Sci Sports Exerc. 2002;34(5):S5.
5. Bonacci J, Hall M, Fox A, et al. The influence of cadence and shoes on patellofemoral joint kinetics in runners with patellofemoral pain. J Sci Med Sport. 2018;21(6):574–8.
6. Milner CE, Ferber R, Pollard CD, et al. Biomechanical factors associated with tibial stress fracture in female runners. Med Sci Sports Exerc. 2006;38(2):323–8.
7. Gabbett TJ, Nassis GP, Oetter E, et al. The athlete monitoring cycle: a practical guide to interpreting and applying training monitoring data. Br J Sports Med. 2017;51(20):1451–2.
8. Bourdon PC, Cardinale M, Murray A, et al. Monitoring athlete training loads: consensus statement. Int J Sports Physiol Perform. 2017;12(Suppl 2):S2161–70.
9. Willy RW. Innovations and pitfalls in the use of wearable devices in the prevention and rehabilitation of running related injuries. Phys Ther Sport. 2018;29:26–33.
10. DeJong Lempke AF, Hart JM, Hryvniak DJ, et al. Prospective running assessments among division I cross-country athletes. Phys Ther Sport. 2022;55:37–45.
11. DeJong Lempke AF, Hart JM, Hryvniak DJ, et al. Use of wearable sensors to identify biomechanical alterations in runners with exercise-related lower leg pain. J Biomech. 2021;126:110646.
12. Koldenhoven RM, Virostek A, DeJong AF, et al. Increased contact time and strength deficits in runners with exercise-related lower leg pain. J Athl Train. 2020;55(12):1247–54.
13. Kiernan D, Hawkins DA, Manoukian MAC, et al. Accelerometer-based prediction of running injury in National Collegiate Athletic Association track athletes. J Biomech. 2018;73:201–9.
14. Koldenhoven RM, Hertel J. Validation of a wearable sensor for measuring running biomechanics. Digit Biomark. 2018;2:74–8.
15. Napier C, Willy RW, Hannigan BC, et al. The effect of footwear, running speed, and location on the validity of two commercially available inertial measurement units during running. Front Sports Act Living. 2021;3:643385.
16. Brayne L, Barnes A, Heller B, Wheat J. Using a wireless consumer accelerometer to measure tibial acceleration during running: agreement with a skin-mounted sensor. Sports Eng. 2018;21(4):487–91.
17. Govus AD, Coutts A, Duffield R, et al. Relationship between pretraining subjective wellness measures, player load, and rating-of-perceived-exertion training load in American college football. Int J Sports Physiol Perform. 2018;13(1):95–101.
18. Gastin PB, Meyer D, Robinson D. Perceptions of wellness to monitor adaptive responses to training and competition in elite Australian football. J Strength Cond Res. 2013;27(9):2518–26.
19. Borg G. Borg’s Perceived Exertion and Pain Scales. Champaign (IL): Human Kinetics; 1998. 104 p.
20. Haddad M, Stylianides G, Djaoui L, et al. Session-RPE method for training load monitoring: validity, ecological usefulness, and influencing factors. Front Neurosci. 2017;11:612.
21. Nelson EO, Ryan M, AufderHeide E, Heiderscheit B. Development of the University of Wisconsin Running Injury and Recovery Index. J Orthop Sports Phys Ther. 2019;49(10):751–60.
22. DeJong AF, Hertel J. Outdoor running activities captured using wearable sensors in adult competitive runners. Int J Athl Ther Train. 2020;25(2):76–85.
23. Zadpoor AA, Nikooyan AA. The relationship between lower-extremity stress fractures and the ground reaction force: a systematic review. Clin Biomech (Bristol, Avon). 2011;26(1):23–8.
24. Bennell K, Crossley K, Jayarajan J, et al. Ground reaction forces and bone parameters in females with tibial stress fracture. Med Sci Sports Exerc. 2004;36(3):397–404.
25. de Oliveira Silva D, Briani R, Pazzinatto M, et al. Vertical ground reaction forces are associated with pain and self-reported functional status in recreational athletes with patellofemoral pain. J Appl Biomech. 2015;31(6):409–14.
26. Becker J, James S, Wayner R, et al. Biomechanical factors associated with Achilles tendinopathy and medial tibial stress syndrome in runners. Am J Sports Med. 2017;45(11):2614–21.
27. Pohl MB, Mullineaux DR, Milner CE, et al. Biomechanical predictors of retrospective tibial stress fractures in runners. J Biomech. 2008;41(6):1160–5.
28. Yong JR, Silder A, Montgomery KL, et al. Acute changes in foot strike pattern and cadence affect running parameters associated with tibial stress fractures. J Biomech. 2018;76:1–7.
29. Heiderscheit BC, Chumanov ES, Michalski MP, et al. Effects of step rate manipulation on joint mechanics during running. Med Sci Sports Exerc. 2011;43(2):296–302.
30. Schubert AG, Kempf J, Heiderscheit BC. Influence of stride frequency and length on running mechanics: a systematic review. Sports Health. 2014;6(3):210–7.
31. Edwards WB, Taylor D, Rudolphi TJ, et al. Effects of stride length and running mileage on a probabilistic stress fracture model. Med Sci Sports Exerc. 2009;41(12):2177–84.
32. Willy RW, Buchenic L, Rogacki K, et al. In-field gait retraining and mobile monitoring to address running biomechanics associated with tibial stress fracture. Scand J Med Sci Sports. 2016;26(2):197–205.
33. Impellizzeri FM, Menaspà P, Coutts AJ, et al. Training load and its role in injury prevention, part I: back to the future. J Athl Train. 2020;55(9):885–92.
34. Campbell PG, Stewart IB, Sirotic AC, et al. Analysing the predictive capacity and dose–response of wellness in load monitoring. J Sports Sci. 2021;39(12):1339–47.

Supplemental Digital Content

Copyright © 2022 by the American College of Sports Medicine