ARTICLE IN PRESS

Available online at www.sciencedirect.com

Journal of Sport and Health Science xxx (2022) xxx-xxx

1 Review 53
2 54
3 Biomechanics associated with tibial stress fracture in runners: A systematic 55
4 56
5 review and meta-analysis 57
6 58
7
Q1 1X XClare
XD E. Milner D2X X *, D3X XEric Foch D4X X , D5X XJoseph M. Gonzales D6X X , D7X XDrew Petersen D8X X
a, b a a
59
a
8
Department of Physical Therapy and Rehabilitation Sciences, Drexel University, Philadelphia, PA 19102, USA 60
b
Department of Health Sciences, Central Washington University, Ellensburg, WA 98926, USA
9 61
Received 16 June 2022; revised 7 October 2022; accepted 4 November 2022
10 62
Available online xxx
11 63
12 2095-2546/Ó 2022 Published by Elsevier B.V. on behalf of Shanghai University of Sport. This is an open access article under the CC BY-NC-ND license 64
13 (http://creativecommons.org/licenses/by-nc-nd/4.0/) 65
14 66
15 Abstract 67
16 68
Background: Tibial stress fracture (TSF) is an overuse running injury with a long recovery period. While many running studies refer to bio-
17 69
mechanical risk factors for TSF, only a few have compared biomechanics in runners with TSF to controls. The aim of this systematic review and
18 70
meta-analysis was to evaluate biomechanics in runners with TSF compared to controls.
19 71
Methods: Electronic databases PubMed, Web of Science, SPORTDiscus, Scopus, Cochrane, and CINAHL were searched. Risk of bias was
20 assessed and meta-analysis conducted for variables reported in 3 or more studies. 72
21 Results: The search retrieved 359 unique records, but only the 14 that compared runners with TSF to controls were included in the review. Most 73
22 studies were retrospective, 2 were prospective, and most had a small sample size (5 30 per group). Many variables were not significantly differ- 74
23 ent between groups. Meta-analysis of peak impact, active, and braking ground reaction forces found no significant differences between groups. 75
24 Individual studies found larger tibial peak anterior tensile stress, peak posterior compressive stress, peak axial acceleration, peak rearfoot ever- 76
25 sion and hip adduction in the TSF group. 77
26
Conclusion: Meta-analysis indicated that discrete ground reaction force variables were not statistically significantly different in runners with TSF 78
compared to controls. In individual included studies, many biomechanical variables were not statistically significantly different between groups.
27 79
However, many were reported by only a single study, and sample sizes were small. We encourage additional studies with larger sample sizes of
28 80
runners with TSF and controls and adequate statistical power to confirm or refute these findings.
29 81
30 Keywords: Bone stress injury; Gait; Kinematics; Kinetics; Tibial acceleration 82
31 83
32 84
33 85
34 1. Introduction capacity for repair and remodeling.3 It is also a serious injury 86
35 with a typical recovery period of up to 8 weeks.4 Furthermore, 87
Running is a popular form of exercise with many health
36 runners are 5 times more likely to experience a recurrence of 88
benefits, but it is also associated with a high rate of overuse
37 stress fracture after the initial injury episode, pointing to an 89
injury, ranging from 19% to 80%.1 Overuse injuries result in
38 underlying factor that is not resolved during rehabilitation 90
time lost from running, which can impact health, well-being,
39 treatments.5 Thus, efforts to reduce the risk of TSF in runners 91
and fitness or competition goals. While many factors both
40 are needed to break the cycle of recurrent and long-lasting 92
internal and external to the body may contribute to overuse
41 periods of injury. Given the frequency and severity of TSF, 93
injury,2 running biomechanics is a readily modified factor and,
42 running biomechanics have been a target of TSF research. 94
therefore, a common target for injury prevention efforts.
43 We have observed that many studies refer to biomechanical 95
Tibial stress fracture (TSF) is a running injury caused by
44 risk factors for TSF when interpreting findings on healthy run- 96
repeated mechanical loading leading to bone strain that creates
45 ners, but only a few have compared biomechanics in runners 97
microcracks at a rate that accumulates beyond the bone’s
46 with TSF to controls. Rather, many studies report biomechan- 98
47 ics of healthy runners only and do not include a TSF group. 99
48 Peer review under responsibility of Shanghai University of Sport. Additionally, existing systematic reviews evaluating the litera- 100
49 * Corresponding author. ture on running biomechanics and injury have only considered 101
50 E-mail address: [email protected] (C.E. Milner). 102
51 103
https://doi.org/10.1016/j.jshs.2022.12.002
52 104

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
2 C.E. Milner et al.
6,7
105 vertical ground reaction force variables. Thus, there is a Biomechanical data for the comparison of TSF and control 162
106 need to systematically review the literature that compares the groups were extracted from all included articles by a single 163
107 biomechanics of runners with TSF to controls to determine the reviewer (JG) and put into a spreadsheet. Group means, stan- 164
108 strength of current evidence for biomechanical differences and dard deviations, and sample sizes were extracted, plus effect 165
109 to identify gaps in the literature as well as areas where further size, if reported, and p values for group comparisons. A second 166
110 research is needed. Therefore, the aim of this systematic reviewer (DP) confirmed the extracted data’s fidelity with the 167
111 review and meta-analysis was to examine biomechanics in original articles. Any discrepancies between reviewers were 168
112 runners with TSF compared to controls by evaluating and syn- resolved by discussion and review of the articles with the third 169
113 thesizing the peer-reviewed literature. reviewer (CM). All quantitative biomechanical variables 170
114 reported in the articles were extracted. When an outcome vari- 171
115 2. Methods able was reported more than once from the same large research 172
116 study, only the findings from the report with the largest sample 173
117 2.1. Literature search size were included to avoid over-representing the study in this 174
118 We conducted a systematic review of the published peer- review. Cohen’s d effect sizes12 were calculated when group 175
119 reviewed literature reporting running biomechanics associated means and standard deviations were provided in studies that 176
120 with TSF. The review and protocol were not registered but did not report effect size. Data were compiled into tables for 177
121 were conducted according to published Preferred Reporting presentation of results. 178
122 Items for Systematic reviews and Meta-Analyses (PRISMA) 179
123 guidelines.8 10 A research question was developed according 180
124 to the 3 elements detailed by Booth et al.11 (who: runnersD9X X; 2.3. Risk of bias assessment
181
125 what: TSF;D10X X how: biomechanics): Which biomechanical param- 182
126 eters distinguish runners with TSF from runners without TSF? The risk of bias in included articles was assessed by 2 183
127 A search strategy was developed with the assistance of a librar- reviewers (JG and DP) using 2 tools. Included studies were 184
128 ian to define search terms for the study participants (runners), evaluated according to the Joanna Briggs Institute checklist 185
129 target condition (TSF), and outcome measures of interest (bio- for analytical cross-sectional studies.13 The checklist was 186
130 mechanical variables measured during running). The elec- modified by removing an item about exposure, leaving a total 187
131 tronic databases PubMed, Web of Science, SPORTDiscus, of 7 items for appraisal. Articles were scored 0 7 with 1 point 188
132 Scopus, Cochrane, and CINAHL were searched for published given for each “yes” answer to checklist questions about study 189
133 peer-reviewed articles and abstracts from all years up to May methods and statistical analysis. Answers of “no” or “unclear” 190
134 2021. Literature review articles and articles in languages other were given 0 points. Included studies were also evaluated for 191
135 than English were excluded. The complete search strategy for the quality of study design, reporting of results, and risk of 192
136 all databases is detailed in Supplementary Table 1. Addition- bias using the AXIS tool for cross-sectional studies.14 Any dis- 193
137 ally, a hand search of the reference lists of review articles iden- crepancies between reviewers were resolved by discussion and 194
138 tified during the search was conducted. review of the article with the third reviewer (CM). An overall 195
139 Two reviewers (CM and DP) independently screened items risk of bias was determined for each study based on the collec- 196
140 for inclusion in the review in 3 rounds. First, articles that could tive findings of these evaluations. 197
141 be excluded based on title were excluded from further review. 198
142 Second, articles were excluded based on their abstract. Third, 199
143 the full text of all remaining articles was retrieved, then 200
2.4. Meta-analysis
144 articles were excluded following review of the full text by 201
145 both reviewers. The final lists of potential articles for inclusion To be included in the meta-analysis, a variable must have 202
146 from each reviewer were compared. Disagreements in article been reported for both TSF and control groups in 3 or more 203
147 selection were resolved by discussion and joint review of the studies. Mean, standard deviation, and group sample size for 204
148 full text until consensus was reached. each variable were entered into the software Review Manager 205
149 Version 5.41. (RevMan, Copenhagen, Denmark). Separate 206
150 meta-analyses were performed for each continuous variable. 207
2.2. Data extraction
151 Group mean differences were analyzed via an inverse variance 208
152 Study details were extracted into a spreadsheet indepen- fixed-effect model.15 This statistical model weights the effect 209
153 dently by 2 reviewers (DP and JG). Details extracted included of each study by the inverse of the variance from each study 210
154 country, year of publication, group definitions and inclusion/ included in the meta-analysis. Group mean differences were 211
155 exclusion criteria, participant characteristics (weekly mileage, considered different from 0 if the overall effect was p < 0.05. 212
156 sex, running level, foot strike pattern of participants), sample Ninety-five percent confidence intervals were also computed 213
157 sizes, experimental location, running velocity, footwear during for the mean differences within each study. The test statistics 214
158 testing, and primary data analysis. Extracted details for each x2 (with corresponding p value) and heterogeneity (I2) were 215
159 study were then compared, and discrepancies between used to describe the amount of heterogeneity across studies in 216
160 reviewers were resolved by discussion and review of the article each meta-analysis.15 I2 was considered low (25%), moderate 217
161 with the third reviewer (CM). (50%), and high (75%).15 218

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
Running biomechanics and tibial stress fracture 3

219 3. Results some variation in the definition of the control group. Runners 276
220 in both groups were healthy and free of injury at the time of 277
3.1. Study selection
221 data collection in most studies. One study included currently 278
222 The initial search retrieved 684 records, resulting in 359 injured runners,22 and 2 others were prospective.23,24 Most 279
223 unique records when duplicates were removed (Fig. 1). Fol- studies were conducted indoors, looked at overground running, 280
224 lowing evaluation of title and abstract, 337 items were and reflected the traditional gait analysis laboratory setting. 281
225 excluded, and 28 full texts were retrieved for assessment, with One study was conducted on an indoor running track with the 282
226 14 items (12 research articles and 2 conference abstracts) runner making contact with the force platform once per lap,25 283
227 retained for inclusion in the review. and another was conducted on an instrumented treadmill.22 284
228 When reported, running velocity was fixed and ranged from 285
229 3.6 m/s to 4.0 m/s. In the treadmill study, running velocity was 286
230 3.2. Study characteristics self-selected and averaged 2.60 m/s in runners with current 287
231
The 14 included publications were published between 1993 TSF and 2.65 m/s in controls.22 288
232
and 2020 in Australia, Canada, the UK, and the USA (Table 1). Discrete biomechanical outcome variables were measured 289
233
Of these 14, 6 were from the same larger research study report- and/or calculated in all studies. Direct measurements were 290
234
ing runners with a history of TSF.16 21 Sample sizes ranged made of lower extremity kinematics, tibial acceleration, and 291
235
from 5D1X X to 30 participants per group, and a priori power analy- ground reaction forces during running. Lower extremity kinet- 292
236
sis for sample size justification was provided in 6 studies ics were calculated via inverse dynamics. Bone stress variables 293
237
(Table 1). All studies compared a group of runners with cur- were modeled from 3-dimensional gait analysis combined with 294
238
rent, future, or a history of TSF to a healthy control group. bone parameters determined from tibial x-rays. Following tra- 295
239
There was some variation in how the groups were defined ditional gait analysis methods, peak magnitudes for variables 296
240
according to study inclusion and exclusion criteria. Occurrence of interest or magnitude at defined time points in the stride 297
241
of TSF was confirmed by a medical professional and/or con- cycle were extracted from the time series data. Magnitudes 298
242
firmed by imaging in all but a single study. There was also were averaged across multiple trials per participant, and group 299
243 differences analyzed to identify statistically significant differ- 300
244 ences. 301
245 302
246 3.3. Risk of bias in included studies 303
247 304
The majority (12/14) of articles scored 7/7 for methodologi-
248 305
cal quality according to the Joanna Briggs Institute checklist,
249 306
indicating a low risk of bias. Two conference abstracts scored
250 307
2/7 and 3/7, respectively, indicating greater risk of bias (Sup-
251 308
plementary Table 2). The AXIS appraisal tool mirrored these
252 309
findings, with the 2 abstracts being of lower methodological
253 310
quality due to their brevity than the majority of studies, which
254 311
were high quality (Supplementary Table 3). Thus, except for
255 312
the 2 abstracts, all included studies were considered to have a
256 313
low risk of bias.
257 314
258 315
3.4. Findings of included studies
259 316
260 3.4.1. Meta-analysis results 317
261 Due to the greater risk of bias, data reported in either of the 318
262 2 abstracts were not included in the meta-analysis. Therefore, 319
263 meta-analyses were conducted for 3 variables: peak vertical 320
264 impact force, peak vertical active force, and peak braking 321
265 force. The results present insufficient evidence to reject the 322
266 null hypothesis of no difference between groups for any of the 323
267 ground reaction force variables included in the meta-analysis 324
268 of runners with previous TSF vs. controls (p > 0.05). Specifi- 325
269 cally, the meta-analysis for peak vertical impact force included 326
270 122 runners and had a p-value of 0.92 with moderate heteroge- 327
271 neity (I2 = 57%; Fig. 2). The meta-analysis for peak vertical 328
272 active force, included 170 runners and had a p-value of 0.36 329
273 with low heterogeneity (I2 = 0%; Fig. 3). Lastly, the meta-anal- 330
274 Fig. 1. Preferred Reporting Items for Systematic reviews and Meta-Analyses ysis for peak braking force included 170 runners and had a p 331
275 (PRISMA) 202010 flow diagram. value of 0.53 with low heterogeneity (I2 = 0%; Fig. 4.) 332

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
4 C.E. Milner et al.

333 Table 1 390

334 Description of included studies. 391
335 Study and year Participants and sample size A priori power Weekly mileage Type of runner Experimental Velocity (m/s) and 392
336 analysis? (mean § SD) and foot strike location footwear (if 393
337 pattern (if reported) 394
338 reported) 395
339 Grimston et al., (1993)25 5 retrospective TSF (defini- No NR NR Indoor track, early NR 396
340 tion NR), 5 control (no his- and late stages of 397
tory of stress fracture) 45-min run.
341 398
Crossley et al., (1999)27 23 currently healthy, retro- No, post hoc 95% NR Male rearfoot 30 m, indoor, 4.0 § 10%; own
342 spective TSF (physician diag- power to detect strike running over-ground running shoes 399
343 nosis, confirmed by imaging), difference of 10% athletes 400
344 23 currently healthy control between groups 401
345 (no history of stress fracture) 402
Bennell et al., (2004)26 13 currently healthy, retro- No, post hoc sam- TSF: 33.7 § 20.1; Female running 30 meters, indoor, 4.0 § 0.4; own
346 403
spective TSF (physician diag- ple size sufficient control: 30.3 § athletes over-ground running shoes
347 nosis, confirmed by imaging), to detect 10% dif- 21.6 404
348 22 currently healthy control ference between 405
349 (no history of stress fracture) groups 406
350 Davis et al., (2004)23 5 prospective TSF (docu- No NR Competitive 25 m, indoor, 3.8 407
mented TSF or tibial stress women runners over-ground in
351 408
reaction), 5 control (defini- laboratory
352 tion NR) 409
353 Milner et al., (2006)17 20 currently healthy, retro- Yes, 20 per group TSF: 28.6 § 6.8; Recreational and 23 m, indoor, 3.7 § 5%; neutral 410
354 spective TSF (medical pro- to detect 15% dif- control: 29.2 § competitive rear- over-ground in running shoes 411
355 fessional diagnosis, ference, alpha 9.9 foot strike female laboratory 412
confirmed by imaging), 20 0.05, 80% power runners
356 413
currently healthy control (no using preliminary
357 previous lower extremity data 414
358 bony injury) 415
359 Milner et al., (2006)16 25 currently healthy, retro- Yes, 24 per group TSF: 28.6 § 9.3; Recreational and 23 m, indoor, 3.7 § 5%; neutral 416
360 spective TSF (medical pro- for effect size control: 28.6 § competitive rear- over-ground running shoes 417
fessional diagnosis, 0.48, a 0.05, 80% 11.8 foot strike female
361 418
confirmed by imaging), 25 power using cited runners
362 419
currently healthy control (no preliminary data
363 previous lower extremity 420
364 fractures) 421
365 Zifchock et al., (2006)21 24 currently healthy, retro- Yes, 24 for a 0.05, TSF: 26.8 § 10.8; Female rearfoot 25 m, indoor, 3.7 § 5%; neutral 422
spective TSF (history of 1 or 80% power using Control: 28.8 § strike runners over-ground running shoes
366 423
more unilateral TSF, con- data from 11.8
367 424
firmed by imaging), 25 cur- literature
368 rently healthy control (never 425
369 sustained a running injury) 426
370 Milner et al., (2007)18 23 currently healthy, retro- Yes, 19 per group TSF: 29.2 § 8.7; Recreational rear- 23 m, indoor, 3.7 § 5%; neutral 427
spective TSF (medical pro- to detect 15% dif- control: 28.6 § foot strike female over-ground in running shoes
371 428
fessional diagnosis, ference, a 0.05, 9.3 runners laboratory
372 429
confirmed by imaging), 23 80% power using
373 currently healthy control (no preliminary data 430
374 previous lower extremity 431
375 bony injury) 432
Creaby et al., (2008)28 10 currently healthy, retro- No NR Male rearfoot 20 m, indoor, 3.6 § 5%; above
376 433
spective TSF (confirmed by strike military over-ground ankle combat
377 434
principal medical officer recruits assault boots
378 using an imaging algorithm), 435
379 20 currently healthy control 436
380 (no history of lower limb 437
injury sustained during
381 438
training)
382 439
Pohl et al., (2008)20 30 currently healthy, retro- No TSF: 25.5 § 6.8; Female rearfoot 25 m, indoor, 3.7 § 5%; neutral
383 spective TSF (confirmed by a control: 24.2 § strike runners over-ground running shoes 440
384 medical doctor), 30 currently 8.7 441
385 healthy control (no previous 442
lower extremity bony injury)
386 443
387 (continued on next page) 444
388 445
389 446

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
Running biomechanics and tibial stress fracture 5

447 Table 1 (Continued) 504

448 505
Study and year Participants and sample size A priori power Weekly mileage Type of runner Experimental Velocity (m/s) and
449 506
analysis? (mean § SD) and foot strike location footwear (if
450 pattern (if reported) 507
451 reported) 508
452 Milner et al., (2010) 19
29 currently healthy, retro- Yes, 29 per group TSF: 28.6 § 13.0; Female rearfoot Indoor, over- 3.7 § 5%; neutral 509
453 spective TSF (medical pro- to detect differ- Control: 26.7 § strike running ground in running shoes 510
454 fessional diagnosis, ence of 1 standard 7.5 athletes laboratory 511
455
confirmed by imaging), 29 deviation, a 0.05, 512
currently healthy control (no 80% power using
456 513
previous lower extremity preliminary data
457 bony injury) 514
458 Meardon et al., (2015)29 23 currently healthy, retro- Yes, 23 per group TSF: 18.2 § 10.7; Male or female 25 m, indoor, 3.7; neutral run- 515
459 spective TSF (physician diag- to detect 15% dif- control: 17.8 § runners over-ground ning shoes 516
460
nosis, confirmed by imaging), ference, alpha 10.6 517
23 currently healthy control 0.05, 80% power
461 518
(no history of stress fracture) using data from
462 literature 519
463 Nunns et al., (2016)24 10 prospective TSF (con- No NR Military recruits 15 m, indoor, 3.6 § 5%; 520
464 firmed by imaging), 150 cur- over-ground barefoot 521
465
rently healthy control 522
(recruits who completed
466 523
training without injury)
467 Johnson et al., (2020) 22
23 currently injured TSF (tib- No NR Male or female Instrumented Self-selected 524
468 ial bone stress injury, con- rearfoot strike treadmill velocity. TSF 525
469 firmed by imaging, pain 2/ runners (2.65 § 0.24), 526
470
10), 65 currently healthy con- Controls (2.60 § 527
trol (injury-free for previous 0.22); lab shoes
471 528
3 months) matched to habit-
472 ual footwear 529
473 530
Abbreviations: NR = not reported; SD = standard deviation; TSF = tibial stress fracture.
474 531
475 532
476 3.4.2. Systematic review results prospective TSF group compared to controls reported in 533
477 A total of 25 kinematic and kinetic variables were reported, another abstract,23 but it was not statistically significantly dif- 534
478 with many having no effects and insufficient evidence to reject ferent in other studies.20,22 Peak adduction free moment was 535
479 the null hypothesis of no significant differences between significantly larger in the TSF group and with a large effect 536
480 groups (Table 2). However, some significant differences (p < compared to controls in one study16 but not statistically signifi- 537
481 0.05) were noted as follows: Peak eversion angle was signifi- cantly different and with no effect in another.28 Free moment 538
482 cantly larger with a moderate effect in the TSF group com- at peak braking force was significantly larger in the TSF group 539
483 pared to controls.19 At the hip, the peak adduction angle was and with a moderate effect compared to controls.16 540
484 larger in the TSF group, with a large effect compared to con- Tibial bone stress during running was reported in a recent 541
485 trols.19 Early D12X Xstance sagittal plane knee joint stiffness was sig- study by Meardon et al.29 (Supplementary Table 4). Peak ante- 542
486 nificantly larger with a large effect18 in the TSF group rior tensile stress and peak posterior compressive stress at the 543
487 compared to controls. Additionally, tibial rotation range of distal third of the tibia were both significantly larger in the 544
488 motion was smaller, with a moderate effect (compared to con- TSF group and with moderate effects compared to controls. 545
489 trols) during barefoot running in recruits who went on to sus- 546
490 tain TSF.24 547
4. Discussion
491 A total of 38 different ground reaction force variables were 548
492 reported, with most having no to small effects and insufficient The purpose of this systematic review and meta-analysis 549
493 evidence to reject the null hypothesis of no significant differ- was to examine biomechanics in runners with TSF compared 550
494 ences between groups (Table 3). Peak vertical impact force to controls by evaluating and synthesizing the peer-reviewed 551
495 was significantly smaller in the TSF group compared to con- literature. Overall, 359 relevant publications were found, but 552
496 trols in ant abstract25 but not statistically significantly different only 14 compared runners with TSF to a control group. Risk 553
497 in other studies.17,26,27 Similarly, in the same abstract peak of bias was low, except for the 2 abstracts, which lacked meth- 554
498 propulsive force was significantly smaller in the TSF group,25 odological detail due to their short length. All 359 publications 555
499 but it was not statistically significantly different in another identified in the search included some combination of the 556
500 study.26 It should be noted that the abstract25 did not indicate tibia/tibial and stress fracture/bony injury terms as indicated in 557
501 whether standard deviation or standard error of the mean was Table 1. However, the vast majority of those studies did not 558
502 reported; thus, effect sizes were not calculated here. Vertical include a group of runners with TSF, despite appearing in a 559
503 instantaneous loading rate was significantly larger in a small search specifically including TSF in the terms. While there is 560

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
6 C.E. Milner et al.

561 618
562 619
563 620
564 621
565 622
566 623
567 624
568 625
569
Fig. 2. Forest plot of vertical impact peak during running showing no difference between groups. Normalized to body weight. 95%CI = 95% confidence interval; 626
IV = inverse variance; TSF = tibial stress fracture.
570 627
571 628
572 629
573 630
574 631
575 632
576 633
577 634
578 635
579 Fig. 3. Forest plot of peak vertical active force during running showing no difference between groups. Normalized to body weight. 95%CI = 95% confidence inter- 636
580 val; IV = inverse variance; TSF = tibial stress fracture. 637
581 638
582 639
583 640
584 641
585 642
586 643
587 644
588 645
589 646
590
Fig. 4. Forest plot of peak braking force during running showing no difference between groups. Normalized to body weight. 95%CI = 95% confidence interval; 647
IV = inverse variance; TSF = tibial stress fracture.
591 648
592 649
clearly a great deal of interest in understanding the biomechan- differences. Thus, the systematic review findings for variables
593 650
ics associated with TSF in runners, few studies have ade- reported by only 1 or 2 studies should be considered prelimi-
594 651
quately addressed this by including runners with the injury. nary and suggestive of variables that may be further investi-
595 652
gated in the effort to understand biomechanical differences
596 653
4.1. Sample size of included studies between runners with TSF and controls.
597 654
598 Overall, the sample sizes of these studies were rather small. 655
According to G*Power,30 for an independent samples t test 4.2. Variables for further investigation
599 656
600 with 80% power and p < 0.05, a sample size of 26 participants Variables from the present studies with moderate or larger 657
601 per group would be needed to detect a significant difference effect sizes (and pD> 13X X 0.05) may also be considered worthy of 658
602 between groups for a large effect size, and 64 per group to further investigation. Thus, dependent variables to be consid- 659
603 detect a significant difference for a moderate effect size. Only ered includeD14X X peak hip adduction angle,20 peak rearfoot ever- 660
604 2 studies included more than 26 participants per group,19,20 sion,19 tibial internal rotation and rearfoot eversion at impact 661
605 and none had more than 30 per group. Thus, 10 of the 14 peak,19 sagittal plane average knee stiffness,17 vertical impact 662
606 reports were underpowered to detect large effects, and all were peak,17 average braking ground reaction force,26 absolute peak 663
607 underpowered to detect moderate effects. We acknowledge free moment,20 frontal and sagittal plane vertical ground reac- 664
608 that increasing the sample size in studies of runners with TSF tion force active peaks,28 angle of the frontal plane vertical 665
609 will take more resources, which may be a practical limitation ground reaction force vector at active peak,28 and peak heel 666
610 for many researchers seeking to investigate the biomechanics pressure.24 Overall, the current body of literature comparing 667
611 of TSF in runners. These practical limitations are likely major runners with TSF to controls identified several biomechanical 668
612 contributing factors to the small sample sizes in many of the variables that may be larger in the TSF group and, therefore, 669
613 included studies. Differences of 15% or 1 standard deviation appropriate for further investigation. 670
614 were used to determine sample size in several included studies. Of the 67 dependent variables identified, only 3 were 671
615 However, if there are important differences between TSF and reported in 3 or more higher quality studies, and so only these 672
616 control groups that are smaller than this (e.g., moderate effect 3 were included in the meta-analysis. All of these variables 673
617 sizes), they are not likely to be identified as significant were discrete ground reaction force variables—vertical impact, 674

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
Running biomechanics and tibial stress fracture 7

675 Table 2 732

676 Kinematic and kinetic variables in tibial stress fracture (TSF) and control groups. 733
677 Variable Study TSF (mean § SD) Control mean (SD) Effect size p Value 734
678 735
Peak rearfoot eversion angle (˚) Milner et al.19 11.7 § 4.2 9.0 § 3.9 0.66 0.015
679 Impact peak rearfoot eversion angle (˚) Milner et al.19 5.5 § 4.0 3.0 § 4.3 0.61 >0.05 736
680 Knee flexion at footstrike (˚) Milner et al.18 13.7 § 6.0 11.9 § 6.5 0.28 0.348 737
681 Knee flexion excursion (˚) Milner et al.17 33.1 § 5.0 34.8 § 5.2 0.34 0.147 738
682 Early stance knee flexion excursion (˚) Milner et al.18 14.4 § 4.0 16.0 § 5.3 0.36 0.252 739
Peak knee adduction angle (˚) Milner et al.19 1.4 § 4.0 2.2 § 5.2 0.17 0.505
683 740
Impact peak knee adduction angle (˚) Milner et al.19 ‒1.6 § 3.6 ‒0.2 § 5.1 0.33 >0.05
684 Peak knee internal rotation angle (˚) Milner et al.19 3.9 § 5.2 3.1 § 6.9 0.13 0.633 741
685 Impact peak knee internal rotation angle (˚) Milner et al.19 ‒2.0 § 6.3 ‒4.4 § 6.4 0.37 >0.05 742
686 Shank angle at footstrike (˚) Milner et al.18 12.8 § 3.4 14.1 § 3.3 0.40 0.181 743
687 Peak tibial internal rotation angle (˚) Milner et al.19 ‒9.4 § 5.8 ‒6.7 § 5.6 0.47 0.080 744
Impact peak tibial internal rotation angle (˚) Milner et al.19 ‒7.1 § 4.6 ‒4.2 § 6.4 0.53 >0.05
688 745
Tibial rotation range of motion (˚) Nunns et al.24 6.4 § 4.3 10.3 § 6.0 ‒0.66* 0.05
689 Peak hip adduction angle (˚) Milner et al.19 11.6 § 5.0 8.1 § 3.7 0.80 0.004 746
690 Impact peak hip adduction angle (˚) Milner et al.19 6.3 § 6.6 4.7 § 4.9 0.29 >0.05 747
691 Peak hip internal rotation angle (˚) Milner et al.19 6.6 § 5.0 8.5 § 6.1 0.33 0.222 748
692 Impact peak hip internal rotation angle (˚) Milner et al.19 2.4 § 6.1 3.7 § 6.9 0.20 >0.05 749
Peak axial tibial acceleration (g) Pohl et al.20 6.5 § 3.4 5.5 § 2.5 0.3 NR
693 750
Peak positive tibial acceleration symmetry index (%) Zifchock et al.21 29.0 § 23.3 31.7 § 24.1 ‒0.11* 0.70
694 Sagittal plane average ankle joint stiffness (£10 2)a Milner et al.17 4.31 § 0.59 4.59 § 0.61 ‒0.46 <0.05 751
695 Sagittal plane average knee joint stiffness (£10 2)a Milner et al.17 4.88 § 0.88 4.46 § 0.68 0.54 0.054 752
696 Early stance sagittal plane average knee joint stiffness (£10 2)a Milner et al.18 4.4 § 2.1 3.0 §1.5 0.79 0.015 753
697 Lower extremity stiffness (kN/m) Davis et al.23 9.21 (NR) 9.63 (NR) NR 0.30 754
Vertical stiffness at initial loading (kN/m) Johnson et al.22 77.59 § 22.48 68.32 § 18.91 0.47* >0.05
698 755
Peak heel pressure (N/cm2) Nunns et al.24 20.81 § 6.83 17.69 § 4.75 0.64* 0.06
699 756
a
Joint stiffness is change in joint moment (Nm/(mass in kg*height in m) divided by change in joint angle (˚).
700 * 757
Calculated effect size. Note that when an outcome variable was reported more than once from the same study, only the report with the largest sample size is
701 758
included here.
702 Abbreviations: kN = kilonewton; Nm = Newton-meter; NR = not reported; SD = standard deviation. 759
703 760
704 active peaks, peak braking force—and were not different in included with variables characterizing running biomechanics 761
705 runners with TSF compared to controls. This is not surprising to help determine which gait patterns reduce tibial load. 762
706 as ground reaction force is a response to the acceleration of the There was a pattern of frontal plane kinematic differences 763
707 center of mass of the whole body. Therefore, observed ground with greater peak rearfoot eversion and hip adduction 764
708 reaction forces cannot be attributed to an individual joint or angles19,20 but smaller tibial rotation range of motion during 765
709 specific body segment, such as the tibia.31 It is feasible that running in the TSF group compared to controls.24 Differences 766
710 many different body segment acceleration configurations in frontal plane alignment and transverse plane motion in run- 767
711 could result in the same magnitude of ground reaction force ners with TSF may alter the distribution of loading on the tibia 768
712 peaks. Bone stress is influenced by muscle forces and joint from that seen in controls.19 Since rearfoot eversion is coupled 769
713 reaction forces in addition to ground reaction forces.29 There- with tibial rotation at the subtalar joint,32 a smaller range of 770
714 fore, discrete ground reaction force variables are likely not tibial rotation with greater peak rearfoot eversion may increase 771
715 sensitive enough to indicate differences between TSF and con- torsional loads within the tibia and contribute to increased 772
716 trol runners. bone stress. Therefore, these differences in lower extremity 773
717 Of the remaining 64 ground reaction force, kinematic, frontal plane alignment may alter the distribution of forces on 774
718 kinetic, and bone stress variables, the majority were reported the tibia, possibly contributing to the risk of injury. 775
719 in only a single study. Most variables reported were not signifi- 776
720 cantly different between groups, with only small effects identi- 777
4.3. Characteristics of study participants
721 fied. However, the study modeling tibial bone stress found 778
722 significant moderate to large effects for higher anterior tensile Per our review criteria, all TSF groups included runners 779
723 and posterior compressive tibial stress in the TSF group com- with TSF. Most studies published as original research articles 780
724 pared to controls.29 These findings suggest that the tibia is were retrospective and included runners who had a previous 781
725 loading differently during running in those with TSF compared TSF from which they had recovered and who were injury-free 782
726 to controls. Approaches that seek to characterize the biome- at the time of testing. However, 1 study did include runners 783
727 chanics of the tibia during running, including bone stress and who were currently injured.22 Furthermore, 2 studies with a 784
728 strain, may prove fruitful in teasing out conditions that small number of runners in the TSF group were 785
729 increase the magnitude of tibial loading. This may help to prospective.23,24 As is typical for overuse running injuries, no 786
730 identify targets for intervention to reduce injury risk during prospective biomechanical studies with a large sample size of 787
731 running.29 These tibial measures may be most insightful when injured runners have been reported. However, given that there 788

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
8 C.E. Milner et al.

789 Table 3 846

790 Ground reaction force variables in tibial stress fracture (TSF) and control groups. 847
791 Variable Study TSF (mean § SD) Control (mean § SD) Effect size p Value 848
792 849
Peak vertical impact force (BW) Grimston et al.25 1.84 2.24 ‒ Significant
793 Crossley et al.27 1.890 § 0.387 1.970 § 0.337 ‒0.22* >0.05 850
794 Bennell et al.26 1.944 § 0.295 2.080 § 0.381 ‒0.39* 0.32 851
795 Milner et al.17 1.84 § 0.21 1.70 § 0.32 0.51 0.057 852
796 Time to peak vertical impact force (s) Crossley et al.27 0.031 § 0.005 0.031 § 0.005 0.00* >0.05 853
Bennell et al.26 0.136 § 0.016 0.132 § 0.023 0.19* 0.65
797 854
Peak vertical active force (BW) Grimston et al.25 2.48 2.68 ‒ Significant
798 Crossley et al.27 2.843 § 0.235 2.856 § 0.189 ‒0.06* >0.05 855
799 Bennell et al.26 2.747 § 0.216 2.786 § 0.247 ‒0.16* 0.47 856
800 Davis et al.23 2.55 (NR) 2.63 (NR) NR 0.15 857
801 Johnson et al.22 2.24 § 0.22 2.28 § 0.22 ‒0.18* NR 858
Time to peak vertical active force (s) Crossley et al.27 0.099 § 0.009 0.097 § 0.011 0.20* >0.05
802 859
Bennell et al.26 0.452 § 0.028 0.451 § 0.047 0.02* 0.94
803 Average vertical force (BW) Bennell et al.26 1.654 § 0.138 1.696 § 0.130 ‒0.32* 0.37 860
804 Peak braking force (BW) Grimston et al.25 0.28 0.35 ‒ Significant 861
805 Crossley et al.27 ‒0.496 § 0.056 ‒0.492 § 0.104 ‒0.05* >0.05 862
806 Bennell et al.26 ‒0.497§ 0.080 ‒0.515 § 0.088 0.21* 0.54 863
Johnson et al.22 0.25 § 0.06 0.23 § 0.12 0.19 NR
807 864
Time to peak braking force (s) Crossley et al.27 0.050 § 0.008 0.051 § 0.013 ‒0.09* >0.05
808 Bennell et al.26 0.211 § 0.063 0.207 § 0.053 0.07* 0.54 865
809 Average braking force (BW) Bennell et al.26 ‒0.232 § 0.031 ‒0.249 § 0.033 0.53* 0.13 866
810 Peak propulsive force (BW) Grimston et al.25 0.37 0.47 ‒ Significant 867
811 Bennell et al.26 0.369 § 0.064 0.380 § 0.039 ‒0.22* 0.55 868
Time to peak propulsive force (s) Bennell et al.26 0.751 § 0.016 0.757 § 0.016 ‒0.38* 0.35
812 869
Average propulsive force (BW) Bennell et al.26 0.213 § 0.033 0.220 § 0.021 ‒0.27* 0.42
813 Peak medial force (BW) Johnson et al.22 0.083 § 0.023 0.093 § 0.068 ‒0.17 NR 870
814 Peak lateral force (BW) Johnson et al.22 0.071 § 0.037 0.065 § 0.041 0.15* NR 871
815 Sagittal plane impact peak force (BW) Creaby et al.28 1.91 § 0.22 1.81 § 0.26 0.38 0.17 872
816 Angle of sagittal plane impact peak vector (˚) Creaby et al.28 ‒5.65 § 3.23 ‒5.84 § 4.67 0.05 OP 873
Sagittal plane active peak force (BW) Creaby et al.28 2.49 § 0.18 2.67 § 0.21 ‒0.85 OP
817 874
Angle of sagittal plane active peak vector (˚) Creaby et al.28 ‒3.25 § 1.63 ‒2.47 § 1.69 ‒0.46 0.12
818 875
Frontal plane impact peak force (BW) Creaby et al.28 1.9 § 0.22 1.8 § 0.26 0.39 0.16
819 Angle of frontal plane impact peak vector (˚) Creaby et al.28 ‒1.30 § 3.83 ‒1.40 § 3.40 0.03 OP 876
820 Frontal plane active peak force (BW) Creaby et al.28 2.49 § 0.19 2.67 § 0.20 ‒0.87 OP 877
821 Angle of frontal plane active peak vector (˚) Creaby et al.28 ‒2.34 § 1.57 ‒1.39 § 1.36 ‒0.64 0.05 878
Vertical instantaneous loading rate (BW/s) Davis et al.23 112.88 (NR) 81.03 (NR) NR 0.04
822 879
Pohl et al.20 88.2 § 24.7 83.8 § 23.2 0.2 NR
823 880
Johnson et al.22 70.78 § 21.55 63.50 § 20.52 0.35* >0.05
824 Vertical average loading rate (BW/s) Davis et al.23 88.20 (NR) 62.91 (NR) NR 0.06 881
825 Pohl et al.20 74.2 § 23.5 66.0 § 22.4 0.4 NR 882
826 Johnson et al.22 61.18 § 19.60 54.37 § 18.25 0.37* NR 883
Braking instantaneous loading rate (BW/s) Milner et al.17 20.35 § 6.17 19.29 § 4.70 0.19 0.272
827 884
Johnson et al.22 8.66 § 2.86 9.02 § 5.73 ‒0.07* NR
828 885
Braking average loading rate (BW/s) Milner et al.17 8.54 § 3.10 8.37 § 2.25 0.07 0.420
829 Medial instantaneous loading rate (BW/s) Johnson et al.22 8.85 § 5.47 8.30 § 4.54 0.11* NR 886
830 Lateral instantaneous loading rate (BW/s) Johnson et al.22 7.82 § 4.12 7.94 § 7.45 ‒0.02* NR 887
831 Peak vertical impact force symmetry index (%) Zifchock et al.21 8.8 § 13.6 12.6 § 10.1 ‒0.32* 0.27 888
Peak vertical active force symmetry index (%) Zifchock et al.21 2.6 § 1.8 3.1 § 2.5 ‒0.23* 0.42
832 889
Peak braking force symmetry index (%) Zifchock et al.21 10.8 § 12.8 11.4 § 8.8 ‒0.05* 0.85
833 890
Peak lateral force symmetry index (%) Zifchock et al.21 36.3 § 42.9 49.8 § 38.3 ‒0.33* 0.25
834 Peak medial force symmetry index (%) Zifchock et al.21 44.8 § 44.8 37.5 § 28.0 0.20* 0.50 891
835 Peak instantaneous vertical loading rate symmetry index (%) Zifchock et al.21 12.6 § 9.4 15.0 § 10.4 ‒0.24* 0.40 892
836 Vertical average loading rate symmetry index (%) Zifchock et al.21 16.5 § 11.7 23.3 § 17.4 ‒0.46* 0.11 893
Peak adduction free moment (£10 3) Milner et al.17 7.7 § 4.7 4.7 § 2.5 0.80 0.004
837 894
Creaby et al.28 6.2 § 2.4 5.7 § 3.1 0.14 0.35
838 895
Free moment at peak braking force (£10 ) 3
Milner et al.17 4.6 § 5.7 1.6 § 3.7 0.62 0.017
839 Absolute peak free moment (£10 3) Pohl et al.20 9.1 § 4.2 6.1 § 2.5 0.9 NR 896
840 Creaby et al.28 9.5 § 2.1 9.3 § 3.2 0.01 0.4 897
841 Net angular impulse (s, £10 4) Milner et al.16 4.5 § 9.9 1.6 § 5.5 0.36 0.105 898
*
842 Calculated effect size. Significant: p value not provided. Note that when an outcome variable was reported more than once from the same study, only the report 899
843 with the largest sample size is included here. 900
Abbreviations: BW = body weight; s = seconds; NR = not reported; OP = opposite of hypothesized direction; TSF = tibial stress fracture; “‒” = effect size not calcu-
844 901
lated due to data uncertainty.
845 902

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
Running biomechanics and tibial stress fracture 9

903 is a high likelihood of recurrence of TSF following the initial included studies but avoided the risk of masking differences 960
904 injury,5 retrospective studies can provide insight into underly- attributable to TSF that may be inconsistent with other stress 961
905 ing factors that may be associated with the injury and future fracture sites, such as the femur or metatarsals. Furthermore, 962
906 injury recurrence. Criteria for inclusion in the control group given the potential limiting effects on meta-analysis of a small 963
907 varied from no history of TSF specifically to no history of any number of included studies and moderate heterogeneity,36 the 964
908 running injury. If there are unique biomechanical features of finding for peak vertical impact force should be interpreted 965
909 running associated with TSF, the control group would, at a with caution. 966
910 minimum, need to only exclude runners with a TSF. Given the 967
911 reported 19% to 79% incidence of injury in runners,1 exclud- 5. Conclusion 968
912 ing runners with any previous running injury from the control 969
The literature reveals an ongoing interest in identifying
913 group greatly reduces the available pool of participants. Thus, 970
approaches for reducing the risk of TSF in runners by examin-
914 to facilitate the inclusion of larger sample sizes, we recom- 971
ing running biomechanics. However, we found only 14 reports
915 mend that future studies include currently healthy runners with 972
(2 of which were abstracts) that compared aspects of running
916 confirmed history of TSF in the injury group and runners with 973
biomechanics between TSF and control groups, and only 1
917 no history of TSF in the control group. 974
that reported tibial stress. Many variables were reported by
918 The majority of studies focused on female runners, likely 975
only a single study. Many variables were not statistically sig-
919 because female runners have a higher incidence of stress frac- 976
nificantly different between the TSF and control groups. Spe-
920 ture.33 However, 3 studies included only male runners,24,27,28 977
cifically, meta-analysis indicated that the discrete ground
921 2 included both men and women,22,29 and 1 did not report.25 978
reaction force variables vertical impact peak, vertical active
922 Since differences in running biomechanics between men and 979
peak, and peak braking force were not statistically significantly
923 women have been reported for some lower extremity bio- 980
different in runners with TSF compared to controls. Sample
924 mechanical variables,34 it is necessary to confirm that differen- 981
sizes were small, so studies may have been underpowered to
925 ces reported in female runners also occur in male runners and 982
detect important differences. We encourage future studies to
926 vice versa. For example, a study conducted in female runners 983
compare runners with previous, current, or prospective TSF to
927 comparing those with TSF to a control group found larger 984
controls with no history of TSF and to use sample sizes of at
928 peak free moment in the TSF group,14 but a study comparing 985
least 26 per group to detect group differences with large effects
929 male runners found no difference in the same variable between 986
and at least 64 to detect moderate effects. This may require
930 groups.28 However, it cannot be determined whether these are 987
multi-center studies to ensure sufficient statistical power.
931 gender differences or simply conflicting study findings. There- 988
While prospective studies are the gold standard, we acknowl-
932 fore, we recommend that future studies include both male and 989
edge that the resources required for these studies are substan-
933 female runners and power the study so that women and men 990
tial. Thus, we also encourage studies comparing runners with a
934 can be treated as separate subgroups during statistical analysis. 991
history of TSF or current TSF to controls to identify biome-
935 Alternatively, if limited resources prohibit this, we suggest 992
chanics associated with TSF.
936 focusing on either female or male runners. 993
937 All but 2 studies reported laboratory gait analysis during 994
Competing interests
938 overground running in short trials of 15 30 m. Recent work 995
939 with precision wearables found significant differences between The authors declare that they have no competing interests. 996
940 the magnitude of biomechanical variables collected in the tra- 997
941 ditional gait analysis laboratory setting and field measures.35 Authors’ contributions 998
942 In particular, tibial acceleration variables were higher when 999
CM conceived the idea and developed the design of the
943 measured during outdoor running compared to running in the 1000
review, reviewed the literature, wrote the manuscript, and
944 laboratory in healthy runners.35 Thus, biomechanical variables 1001
checked data fidelity; JG and DP also reviewed the literature,
945 associated with TSF during laboratory gait analysis cannot be 1002
contributed to table and figure preparation and data extraction,
946 assumed to have the same magnitudes when measured during 1003
checked data fidelity, provided suggestions and revisions to
947 running in the field. Field-based investigations must seek to 1004
the original draft; EF performed all meta-analyses and associ-
948 determine whether the same differences exist between runners 1005
ated figure preparation and provided suggestions and revisions
949 with TSF and controls when they are in the outdoor environ- 1006
to the original draft. All authors have read and approved the
950 ment. 1007
final version of the manuscript and agree with the order of pre-
951 Some limitations of this review should be noted. The search 1008
sentation of the authors.
952 was restricted to research literature published in English and 1009
953 so does not account for studies published in other languages. 1010
Acknowledgments
954 Our focus was on running biomechanics, which may be modi- 1011
955 fiable, and so we did not include bone geometry or other 15XD XThe authors gratefully acknowledge the assistance of librar- 1012
956 unmodifiable anatomical variables. Since our focus was specif- ian Ms Janice Masud-Paul in developing our search strategy. 1013
957 ically on TSF, we excluded studies that placed runners with Financial support was not received for this review. Data 1014
958 stress fracture at other lower extremity locations within the extracted from individual studies are reported in tables and 1015
959 same group as those with TSF. This reduced the number of supplemental table. 1016

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002
 ARTICLE IN PRESS
10 C.E. Milner et al.

1017 Supplementary materials 18. Milner CE, Hamill J, Davis I. Are knee mechanics during early stance 1074
1018 related to tibial stress fracture in runners? Clin Biomech (Bristol, Avon) 1075
Supplementary material associated with this article can be 2007;22:697–703.
1019 1076
found, in the online version, at doi:10.1016/j.jshs.2022.12.002. 19. Milner CE, Hamill J, Davis IS. Distinct hip and rearfoot kinematics in
1020 1077
female runners with a history of tibial stress fracture. J Orthop Sports
1021 Phys Ther 2010;40:59–66. 1078
References
1022 20. Pohl MB, Mullineaux DR, Milner CE, Hamill J, Davis IS. Biomechanical 1079
1023 1. van Gent RN, Siem D, van Middelkoop M, van Os AG, Bierma-Zeinstra predictors of retrospective tibial stress fractures in runners. J Biomech 1080
1024 SM, Koes BW. Incidence and determinants of lower extremity running 2008;41:1160–5. 1081
injuries in long distance runners: A systematic review. Br J Sports Med 21. Zifchock RA, Davis I, Hamill J. Kinetic asymmetry in female runners
1025 1082
2007;41:469–80. with and without retrospective tibial stress fractures. J Biomech
1026 2. Bennell K, Matheson G, Meeuwisse W, Brukner P. Risk factors for stress 2006;39:2792–7. 1083
1027 fractures. Sports Med 1999;28:91–122. 22. Johnson CD, Tenforde AS, Outerleys J, Reilly J, Davis IS. Impact-related 1084
1028 3. Warden SJ, Davis IS, Fredericson M. Management and prevention of bone ground reaction forces are more strongly associated with some running 1085
stress injuries in long-distance runners. J Orthop Sports Phys Ther injuries than others. Am J Sports Med 2020;48:3072–80.
1029 1086
2014;44:749–65. 23. Davis I, Milner C, Hamill J. Does increased loading during running lead to
1030 1087
4. Bennell K, Brukner P. Preventing and managing stress fractures in ath- tibial stress sractures? A prospective study. Med Sci Sports Exerc
1031 letes. Phys Ther Sport 2005;6:171–80. 2004;36:S58. doi:10.1249/00005768-200405001-00271. 1088
1032 5. Wright AA, Taylor JB, Ford KR, Siska L, Smoliga JM. Risk factors asso- 24. Nunns M, House C, Rice H, et al. Four biomechanical and anthropometric 1089
1033 ciated with lower extremity stress fractures in runners: A systematic measures predict tibial stress fracture: A prospective study of 1065 Royal 1090
1034 review with meta-analysis. Br J Sports Med 2015;49:1517–23. Marines. Br J Sports Med 2016;50:1206–10. 1091
6. Zadpoor AA, Nikooyan AA. The relationship between lower-extremity 25. Grimston SK, Nigg BM, Fisher V, Ajemian SV. External loads throughout
1035 1092
stress fractures and the ground reaction force: A systematic review. Clin a 45 minute run in stress fracture and non-stress fracture runners. In: Pro-
1036 Biomech (Bristol, Avon) 2011;26:23–8. ceedings of the International Society of Biomechanics XIV Congress. 1093
1037 7. van der Worp H, Vrielink JW, Bredeweg SW. Do runners who suffer injuries Paris, France; 1993. 1094
1038 have higher vertical ground reaction forces than those who remain injury-free? 26. Bennell K, Crossley K, Jayarajan J, et al. Ground reaction forces and bone 1095
A systematic review and meta-analysis. Br J Sports Med 2016;50:450–7. parameters in females with tibial stress fracture. Med Sci Sports Exerc
1039 1096
8. Ardern CL, B€ uttner F, Andrade R, et al. Implementing the 27 PRISMA 2004;36:397–404.
1040 1097
2020 statement items for systematic reviews in the sport and exercise 27. Crossley K, Bennell KL, Wrigley T, Oakes BW. Ground reaction forces,
1041 medicine, musculoskeletal rehabilitation and sports science fields: The bone characteristics, and tibial stress fracture in male runners. Med Sci 1098
1042 PERSiST (implementing PRISMA in exercise, rehabilitation, sport medi- Sports Exerc 1999;31:1088–93. 1099
1043 cine and sports science) guidance. Br J Sports Med 2022;56:175–95. 28. Creaby MW, Dixon SJ. External frontal plane loads may be associated 1100
1044 9. Page MJ, Moher D, Bossuyt PM, et al. PRISMA 2020 explanation and with tibial stress fracture. Med Sci Sports Exerc 2008;40:1669–74. 1101
elaboration: Updated guidance and exemplars for reporting systematic 29. Meardon SA, Willson JD, Gries SR, Kernozek TW, Derrick TR. Bone
1045 1102
reviews. BMJ 2021;372:n160. doi:10.1136/bmj.n160. stress in runners with tibial stress fracture. Clin Biomech 2015;30:895–
1046 10. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: 902. 1103
1047 An updated guideline for reporting systematic reviews. BMJ 2021;372: 30. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: A flexible statisti- 1104
1048 n71. doi:10.1136/bmj.n71. cal power analysis program for the social, behavioral, and biomedical sci- 1105
1049 11. Booth A, Papaioannou D, Sutton A. Systematic approaches to a successful ences. Behav Res Methods 2007;39:175–91. 1106
literature review. London: Sage; 2012. 31. Winter D. Biomechanics and Motor Control of Human Movement. Hobo-
1050 1107
12. Cohen J. A power primer. Psychol Bull 1992;112:155–9. ken, NJ: John Wiley & Sons, Inc.; 2009.p.121.
1051 13. Moola S, Munn Z, Tufanaru C, et al. Chapter 7: Systematic reviews of eti- 32. Inman VT. The influence of the foot-ankle complex on the proximal skel- 1108
1052 ology and risk. In: Aromataris E, Munn Z, editors. JBI Manual for Evi- etal structures. Artif Limbs 1969;13:59–65. 1109
1053 dence Synthesis. JBI; 2020. doi:10.46658/JBIMES-20-08. Available at; . 33. Wentz L, Liu PY, Haymes E, Ilich JZ. Females have a greater incidence of 1110
https://synthesismanual.jbi.global. stress fractures than males in both military and athletic populations: A sys-
1054 1111
14. Downes MJ, Brennan ML, Williams HC, Dean RS. Development of a crit- temic review. Mil Med 2011;176:420–30.
1055 1112
ical appraisal tool to assess the quality of cross-sectional studies (AXIS). 34. Ferber R, Davis IM. Williams 3rd DS. Gender differences in lower
1056 BMJ Open 2016;6: e011458. doi:10.1136/bmjopen-2016-011458. extremity mechanics during running. Clin Biomech (Bristol, Avon) 1113
1057 15. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsis- 2003;18:350–7. 1114
1058 tency in meta-analyses. BMJ 2003;327:557–60. 35. Milner CE, Hawkins JL, Aubol KG. Tibial acceleration during running is 1115
1059 16. Milner CE, Davis IS, Hamill J. Free moment as a predictor of tibial stress higher in field testing than indoor testing. Med Sci Sports Exerc 1116
fracture in distance runners. J Biomech 2006;39:2819–25. 2020;52:1361–6.
1060 1117
17. Milner CE, Ferber R, Pollard CD, Hamill J, Davis IS. Biomechanical fac- 36. Seide SE, R€over C, Friede T. Likelihood-based random-effects meta-anal-
1061 tors associated with tibial stress fracture in female runners. Med Sci Sports ysis with few studies: Empirical and simulation studies. BMC Med Res 1118
1062 Exerc 2006;38:323–8. Methodol 2019;19:16. doi:10.1186/s12874-018-0618-3. 1119
1063 1120
1064 1121
1065 1122
1066 1123
1067 1124
1068 1125
1069 1126
1070 1127
1071 1128
1072 1129
1073 1130

Please cite this article as: Clare E. Milner et al., Biomechanics associated with tibial stress fracture in runners: A systematic review and meta-analysis, Journal of Sport and Health
Science (2022), https://doi.org/10.1016/j.jshs.2022.12.002