Effects of form-focused training on running biomechanics: A pilot randomized trial in untrained individuals

PM R XXX (2015) 1-9www.pmrjournal.org

Original Research

Effects of Form-Focused Training on Running Biomechanics:A Pilot Randomized Trial in Untrained Individuals

Deepak Kumar, PT, PhD, Kelly McDermott, PhD, Haojun Feng, BA,Veronica Goldman, BA, Anthony Luke, MD, Richard B. Souza, PT, PhD,

Frederick M. Hecht, MD

Abstract

Objective: To investigate the changes in running biomechanics after training in form-focused running using ChiRunning versus not-form focused training and self-directed training in untrained individuals.Design: Pilot study-randomized controlled trial.Setting: Research institution with tertiary care medical center.Participants: Seventeen subjects (9 men, 8 women) with prehypertension.Methods: Twenty-two participants were randomized to 3 study arms but 17 completed the study. The study arms were: (1) group-based Form-Focused running using ChiRunning (enrolled, n ¼ 10; completed, n ¼ 7); 92) group-based conventional running(enrolled, n ¼ 6; completed, n ¼ 4); and (3) self-directed training with educational materials (enrolled, n ¼ 6; completed, n ¼ 6).The training schedule was prescribed for 8 weeks with 4 weeks of follow-up. All subjects completed overground running motionanalyses before and after training. ClinicalTrials.gov identifier for this study is NCT0158718.Outcomes: Ankle, knee, hip joint peak moments, and powers; average vertical loading rate (AVLR); impact peak; cadence; stridelength; strike index; and stride reach. Paired t tests were used to compare differences within groups over time.Results: Form-Focused group reduced their Stride Reach (P ¼ .047) after the training but not the other groups. Form-Focusedgroup showed a close to significant reduction in knee adduction moment (P ¼ .051) and a reduction in the peak ankle eversionmoment (P ¼ .027). Self-Directed group showed an increase in the running speed (P ¼ .056) and increases in ankle and knee jointpowers and moments.Conclusions: There are differences in the changes in running biomechanics between individuals trained in running form thatemphasizes mid-foot strike, greater cadence, and shorter stride compared with those not trained in the these techniques. Thesedifferences may be associated with reduced lower extremity stress in individuals trained in this running form, but more studies areneeded to confirm these findings in larger samples.

Introduction

Running is a popular mode of physical activity in theUnited States, with a large and growing number of in-dividuals who engage in recreational and competitiverunning [1]. Of these runners, 19%-83% may sustain arunning-related injury during their lifetime [2,3]. Com-mon running-related injuries include patellofemoralpain, iliotibial band syndrome, medial tibial stress syn-drome, tibial or metatarsal stress fractures, plantarfasciitis, and Achilles tendinopathy [4,5]. Abnormalvertical loading rates [6,7], external knee adduction

1934-1482/$ - see front matter ª 2015 by the American Academy of Physihttp://dx.doi.org/10.1016/j.pmrj.2015.01.010

moment [8], stride length [9-11], rear-foot strikepattern [12], and low cadence [13] are thought to berelated to greater risk of sustaining running injuries.Hence, programs have been proposed that purport totrain participants in running form that may result in gaitpatterns that reduce the risk of these injuries and pro-mote greater amounts of running [14].

ChiRunning (ChiLiving, Inc., Asheville, NC) is arunning training program based in part on mindfulnessand body awareness principles of Tai chi [15]. It focuseson elements of running form that include a mid-footstrike, shorter stride, and high cadence. Proponents of

cal Medicine and Rehabilitation

2 Effects of Form-Focused Training on Running Biomechanics

this training technique claim that it reduces the risk ofrunning-related injuries and promotes a greater amountof running activity [15]. A recent cross-sectional studyshowed that ChiRunners had lower average verticalloading rates (AVLRs), less knee extensor work, greatercadence, and greater ankle plantar flexor workcompared with individuals running with a traditionalrear-foot strike pattern [16]. However, prospective andrandomized studies investigating changes in runningbiomechanics after training in the ChiRunning techniqueprogram are lacking.

We performed a pilot exploratory study to investigatethe changes in running biomechanics in individualswithout a current running practice during training in arunning form using the ChiRunning approach andcompared them with those receiving more conventionalrunning training. Participants for this study wererecruited for a pilot randomized, controlled trial of theeffects of running on increased blood pressure in in-dividuals with prehypertension. The population withprehypertension was selected because greater than 30%of American adults have prehypertension [17] and anincreasing physical activity is recommended for thispopulation [18]. The aim of this pilot study was to assessfeasibility and gathering preliminary data on key out-comes for a larger randomized trial that is currentlybeing planned.

Figure 1. CONSORT flow diagram of participants. CONS

Methods

Subjects

Entry criteria for the included having a body massindex <35 kg/m2, a mean blood pressure in range (130-150/80-100 mm Hg) at 2 separate in-person visits, andan interest in using exercise to lower blood pressure.Exclusion criteria included having a current runningpractice or a medical condition such as significantosteoarthritis that precluded significant amounts ofrunning. Participants were recruited from the commu-nity by the use of flyers and other print and online ad-vertisements. All participants completed an extensivephone screen with the study coordinator that includedquestions from the Physical Activity Readiness Ques-tionnaire [19] to assess possible risk associated withstarting an exercise program. Callers were ineligibleif they answered yes to any of the 7 items on thePhysical Activity Readiness Questionnaire or providedother information indicating that starting running maynot be appropriate for them. All participants signed aninformed consent form approved by the InstitutionalCommittee on Human Research.

The study flowchart is shown in Figure 1. We enrolled22 participants in the overall trial. All but 1 subject un-derwent gait assessment; the remaining participant was

ORT ¼ Consolidated Standards of Reporting Trials.

3D. Kumar et al. / PM R XXX (2015) 1-9

the last one enrolled and did not have enough time toschedule and complete the baseline gait assessmentbefore the beginning of the trial. The participants wererandomized to 1 of 3 study arms: (1) a group-based, form-focused running (intervention) (enrolled, n ¼ 10; com-pleted, n ¼ 7); (2) a group-based, running training witha coach without focus on form, (active control, attentionmatched) (enrolled, n ¼ 6; completed, n ¼ 4); and (3)self-directed training with educational materials (self-directed control) (enrolled, n ¼ 6; completed, n ¼ 6).

Participants were randomized by the use ofcomputer-generated random blocks of numbers gener-ated by a statistician who was not a member of thestudy staff. The statistician provided study staff withthe randomized numbers in opaque sealed envelopes.The interventions were delivered over the course of 8weeks. All subjects were instructed in a run/walkapproach to running. The training schedule prescribedwalking and running intervals for 12 weeks; however,the postintervention data were collected between

Table 1The 12-week run/walk training schedule

Week Day 1 Day 2

1 Run 1/Walk 3� 5Total 20

Run 1/Walk 2� 6Total 18

2 Run 2/Walk 2� 5Total 20

Run 2/Walk 2� 5Total 20

3 Run 2/Walk 2� 5Total 25

Run 3/Walk 2� 5Total 25

4 Run 4/Walk 2� 4Total 24

Run 4/Walk 1� 5Total: 25

5 Run 5/Walk 2� 4Total 28

Run 5/Walk 1� 5Total 30

6 Run 6/Walk 2� 4Total 32

Run 6/Walk 1� 5Total 35

7 Run 7/Walk 1� 5Total 40

Run 8/Walk 2� 4Total 40

8 Run 9/Walk 2� 4Total 44

Run 9/Walk 1� 4Total 40

9 Run 10/Walk 2� 3Total 36

Run 10/Walk� 3Total 36

10 Run 15/Walk 2� 2Total 34

Run 10/Walk� 3Total 34

11 Run 20/ Walk 2/Run 15Total 37

Run 20/WalkRun 15Total 37

12 Run 25/ Walk 2/Run 10Total 37

Run 25/WalkRun 10Total 37

Understanding the chart: Run 1/Walk 3 � 5 Total 20 ¼ run for 1 minute theinterval 5 times. The total workout lasts 20 minutes.

weeks 8 and 12. The participants were instructed tobeing each workout with a 5-minute walking warm-up,followed by alternating between running and walkingbreaks throughout the workout. Eventually they wereinstructed to progress up to running for 30 minuteswithout any walking breaks. Participants also wereinstructed to cross-train by devoting 1 of the 4 rest daysto using other forms of exercise. The training schedule isshown in Table 1. Compliance was calculated as per-centage of planned running episodes completed fromparticipant training diaries.

Form-Focused Running (Intervention, n ¼ 10)The intervention group attended 4 training workshops

on Sundaymornings at weeks 0 (baseline), 2, 4, and 8. Thefirst workshop at week 0 was 4 hours long and coveredmost of the basic content, whereas the subsequent ses-sions lasted 2 hours and focused on refinement and morespecialized aspects of the technique. The trainingfocused on using body-awareness to help participants

Day 3 Day 4

Run 1/Walk 2� 5Total 20

Rest

Run 2/Walk 2� 5Total 20

Rest

Run 3/Walk 2� 5Total 25

Rest

Run 4/Walk 1� 5Total 25

Rest

Run 5/Walk 1� 5Total 30

Cross-train 30

Run 7/Walk 2� 4Total 36

Cross-train 30

Run 8/Walk 1� 4Total 36

Cross-train 30

Run 10/Walk 2� 3Total 36

Cross-train 30

2 Run 10/Walk 1� 3Total 33

Cross-train 30

2 Run 15/Walk 1� 2Total 32

Cross-train 30

2/ Run 20/Walk 1/Run 15Total 36

Cross-train 30

1/ Run 30 Cross-train 30

n walk for 3 minutes for 4 minutes total. Repeat the 4 minute run/walk

4 Effects of Form-Focused Training on Running Biomechanics

engage with and change their running form. Specificbiomechanical components of training included forwardlean, mid-foot strike, shorter strides, greater cadence,relaxed legs, and core activation. The training was led bya coach certified in the ChiRunning technique using di-dactics, practice drills, and videotaping to provide in-dividuals with detailed feedback on their running form.

Not Form-Focused Running (Active Control)The active control participants attended training

sessions in the afternoons on the same Sundays that theintervention group met for an equivalent amount oftime. The trainings covered running topics unrelatedto form, including goal setting, pace, mileage, warm-up/cool down, stretching, core strengthening, crosstraining, hydration and nutrition, and shoes and gear. Acoach certified by the USA Track and Field and theRoad Runners Club of America led all of the trainingsessions using didactics, practice drills, and group runs.

Self-Directed ControlThe self-directed control participants were provided

with printed educational materials about starting a self-directed run/walk training program. The printed mate-rials provided information on basic running topics otherthan running form. Examples of the topics included goalsetting, pace, mileage, warm-up/cool down, stretching,core strengthening, cross-training, hydration and nutri-tion, and shoes and gear.

Gait Analysis

Participants were invited to complete a 3-dimensional motion analysis assessment of runningbiomechanics at the University of California, San Fran-cisco Human Performance Center, 1-4 weeks beforeweek 0 (baseline) and between weeks 8 and 12 (post-intervention). At each session, the same researcherplaced 9.5-mm spherical retro-reflective markers on thebony landmarks of their bilateral upper and lower ex-tremities to identify joint centers. Participants’ upperextremities, lower extremities, trunk, and head weretracked by the use of clusters of markers. All partici-pants ran overground while 3-dimensional kinematicdata was collected at 250 Hz via a passive 10-cameravideo recording system (Vicon; Vicon Oxford, Oxford,UK). Kinetic data were collected at 1000 Hz from 2 force

Table 2Mean and SD for age, BMI, and the gender distribution for ChiRunning and

Form-Focused Training Active Contr

Age, y 54.3 (7.9) 56.8 (3.9)BMI, kg/m2 27.2 (2.7) 26.2 (4.1)Men/women 4:3 2:2

BMI ¼ body mass index.

platforms embedded in the floor of the laboratory(AMTI, Watertown, MA). A trial was considered usablewhen there was a clean foot-strike on any of the forceplatforms, and the speed was within �5% of an initialrunning pace selected by the participant as “comfort-able.” We collected 4 usable trials on each foot.

The marker coordinate data were low-pass filteredvia a recursive second-order Butterworth filter with acut-off of 6 Hz. The force platform data were high-passfiltered with a recursive second-order Butterworthfilter with a cut-off of 20 Hz. Euler angles (X-Y-Z) usinga right-handed coordinate system and inverse dy-namics were used to calculate kinematic and kineticdata in Visual3D (C-motion, Georgetown, MD). Datafrom left and right were averaged for each individual.Ankle, knee, and hip joint peak external moments andpeak negative and positive joint powers were calcu-lated for all subjects. AVLR, impact peak, stridelength, cadence, and strike index were calculated onthe basis of published methods [20]. Stride reach wasdefined as distance between the center of mass andthe center of pressure at initial contact. All net jointmoments are expressed as external moments andnormalized to each participant’s body weight (BW) andheight (Nm/BW*height).

Data Analysis

Paired t tests were used in each of the 3 groupsto compare the AVLR, Impact Peak, Stride Length,Cadence, Strike Index, Stride Reach, Stride Length,Joint Powers, and Joint Moments pre- and posttraining.All statistical analyses were done using SPSS (SPSS Inc.Chicago, IL) significance level set at P � .05.

Results

Subjects

Age, body mass index, and gender distribution areshown in Table 2. There were no significant differencesin these characteristics between the groups. The sub-jects who were lost to follow-up were not different fromthe rest of the participants in their demographics. Allgroups completed greater than 80% of planned runningepisodes (form-focused ¼ 84%, active control ¼ 98%,self-directed ¼ 83%).

control groups

ol Self-Directed Control P Value

52.3 (4.3) .54226.8 (3.3) .1403:3 c2 ¼ 0.084, P ¼ .959

Figure 2. Loading-related variables for the 3 groups before and after training.

5D. Kumar et al. / PM R XXX (2015) 1-9

Running Gait

There were no significant differences between thegroups at baseline for any of the biomechanical pa-rameters studied.

Loading-Related VariablesThe results are shown in Figure 2. At baseline, the

form-focused running group had a Strike Index of 9.4 �5.2% (rear-foot strike pattern), the active control grouphad a Strike Index of 24 � 26.8% (rear-foot strikepattern), and the self-directed group had a Strike indexof 16.9 � 20% (rear-foot strike pattern). Paired t testsshowed that, posttraining, the form-focused grouptended to increase their Strike Index, the active controlgroup did not change, and the self-directed group ten-ded to reduce their Strike Index (Figure 2); however,none of these changes was significant. Posttraining, theform-focused running group reduced their Stride Reach(P ¼ .047). The active control group also tended toreduce their Stride Reach, but the difference was notsignificant (P ¼ .088), whereas the self-directed groupdid not show a change in their Stride Reach (P ¼ .854).The decrease in Stride Length was not significant for theForm-Focused group (P ¼ .841) and active control groups(P ¼.509). The self-directed group showed an increasein the Stride Length posttraining (P ¼ .051). The form-focused group tended to increase their Cadence(P ¼ .073), whereas the change in Cadence for the other

2 groups was not significant. The decrease in AVLR andImpact Peak for the form-focused group and the in-crease in AVLR and Impact Peak for the other 2 groupswere not statistically significant. The form-focusedrunning group showed a decrease (P ¼ .086) in runningspeed (pretraining: 3.15 � 0.57 m/s, posttraining 2.92 �0.38 m/s). The active control group (pretraining: 2.99 �0.31 m/s, posttraining 3.44 � 0.64 m/s, P ¼ .132) andthe self-directed groups (pretraining: 2.66 � 0.62 m/s,posttraining 2.93 � 0.41 m/s, P ¼ .056) showed an in-crease in running speed.

Joint PowersResults are shown in Figure 3. Paired t tests showed

that the changes in ankle power after training were notsignificant for the form-focused or the active controlgroups. The self-directed group showed an increase inthe peak positive ankle power (P ¼ .028). Similarly, thechanges in knee power were not significant for the form-focused or the active control groups. The self-directedgroup showed an increase in both peak positive (P ¼.040) and peak negative knee power (P ¼ .054). Thechanges in hip powers were not significant in any of the3 groups.

Joint MomentsThe results for hip, knee, and ankle joint moments

are shown Table 3. The changes in hip moments werenot significant in any of the groups. At the knee, the

Figure 3. Peak positive and negative joint powers at the hip (top row), knee (middle row), and ankle (bottom row) for the 3 groups before andafter training.

6 Effects of Form-Focused Training on Running Biomechanics

peak flexion moment (P ¼ .051) and peak internalrotation moment (P ¼ .055) tended to decrease in theself-directed group. The peak knee adduction momentdecreased in the form-focused group (P ¼ .051). Atthe ankle, the form-focused running group showed asignificant reduction in the peak eversion moment(P ¼ .027).

Discussion

The goal of this pilot randomized controlled trialwas to gather preliminary data for a larger randomizedtrial in the future, including optimizing the studydesign and evaluating recruitment, retention, deliveryof the intervention, and participant compliance, etc.The analyses presented here focus on the explorationof the biomechanical changes with training in form-focused running based on ChiRunning compared witha non-form focused training and self-directed trainingto promote greater activity in individuals with pre-hypertension. Our results demonstrate that it isfeasible to train inexperienced individuals in the form-focused running technique, and there are differencesin the changes in running biomechanics betweenindividuals trained in the form-focused running tech-nique compared with those not trained in the form-focused running technique.

The form-focused training used in this study wasdelivered via the use of ChiRunning principles. Thistraining, and other similar running techniques, is pri-marily aimed at promoting a mid-foot strike pattern,increasing cadence, and reducing stride length. Thesechanges are thought to reduce the risk of running-related injuries by reducing the magnitude of loadingapplied to the lower extremities. As can be seen inFigure 2, the changes for the form-focused group for allof these variables was in the expected direction, with adecrease in AVLR, decrease in Impact Peak, an increasein the Strike Index, decrease in Stride Reach and StrideLength, and an increase in Cadence. These findingssupport the utility and feasibility of this training topromote a running gait pattern that may reduce injuryrisk. A number of these changes, however, were notstatistically significant, likely because of the smallsample size. Considering the feasibility nature of thiswork, these data support further evaluation of thistraining in a larger cohort over a longer duration.

We observed a significant reduction in the StrideReach (distance between the center of mass and thecenter of pressure at initial contact) in the form-focused running group with the training. Shorter stridereach (or less overstriding) has been shown to be relatedto lower loads and potentially a lower risk of tibial stressfractures [10,13]. In the recent cross-sectional study onexperienced ChiRunners, the authors did not report

Table 3Joint moments (mean and SD) before and after the intervention for the 3 groups

Pretraining Posttraining

P ValueMean SD Mean SD

HipExtension moment (BW*Ht)Form-focused �0.098 0.04 �0.097 0.03 .935Active control �0.092 0.01 �0.103 0.02 .464Self-directed �0.079 0.04 �0.083 0.03 .572

Flexion moment (BW*Ht)Form-focused 0.043 0.01 0.038 0.01 .306Active control 0.052 0.01 0.061 0.01 .142Self-directed 0.047 0.01 0.051 0.01 .209

Adduction moment (BW*Ht)Form-focused 0.106 0.02 0.106 0.02 .911Active control 0.104 0.01 0.101 0.02 .670Self-directed 0.102 0.01 0.105 0.01 .482

Internal rotation Moment (BW*Ht)Form-focused 0.027 0.01 0.028 0.00 .725Active control 0.022 0.01 0.019 0.01 .665Self-directed 0.019 0.01 0.020 0.01 .866

KneeFlexion moment (BW*Ht)Form-focused �0.154 0.02 �0.154 0.03 .943Active control �0.138 0.03 �0.150 0.02 .528Self-directed �0.128 0.03 �0.152 0.02 .051

Adduction moment (BW*Ht)Form-focused 0.048 0.01 0.044 0.01 .051Active control 0.039 0.01 0.042 0.02 .667Self-directed 0.043 0.02 0.041 0.02 .335

Internal rotation moment (BW*Ht)Form-focused �0.029 0.01 �0.028 0.01 .665Active control �0.025 0.00 �0.026 0.00 .712Self-directed �0.026 0.01 �0.031 0.01 .055

AnkleDorsiflexion moment (BW*Ht)Form-focused 0.131 0.01 0.127 0.01 .273Active control 0.139 0.02 0.138 0.01 .854Self-directed 0.125 0.01 0.124 0.01 .726

Eversion moment (BW*Ht)Form-focused �0.014 0.01 �0.011 0.00 .027*Active control �0.013 0.01 �0.015 0.01 .310Self-directed �0.016 0.01 �0.013 0.00 .272

Internal rotation moment (BW*Ht)Form-focused 0.023 0.02 0.021 0.01 .613Active control 0.025 0.00 0.021 0.01 .532Self-directed 0.018 0.01 0.019 0.01 .798

BW ¼ body weight; Ht ¼ height.* Statistical significance.

7D. Kumar et al. / PM R XXX (2015) 1-9

Stride Reach but speculated that the rear-foot strikershad a more anterior foot position at initial contactcompared with the experienced ChiRunners. We did notsee significant reductions, however, in the AVLR and theImpact peak for the form-focused running group. Thisfinding could be in part attributable to the small samplesize, which limited our ability to detect statisticallysignificant changes in these measures. After training,the mean AVLR for the ChiRunning group was approxi-mately 72 BW/s. Goss et al [16,21] reported an averageAVLR of 43 BW/s in their cohort of experiencedChiRunners. In their study, the average experience withthe ChiRunning form was approximately 30 months.

Furthermore, they recruited individuals who wererunning at least 12 miles/week. In our study we enrolledindividuals without a current running practice. Theobserved difference in AVLR between the studies couldbe related to these differences in the cohorts.

Although the form-focused runners in our studyshowed a more anterior Strike Index posttraining (14% offoot length) compared with pretraining (9% of footlength), as might be expected because it encourages amid-foot strike, the difference was not statisticallysignificant. Less overstriding by the ChiRunners asobserved in our study could be associated with a moreanterior foot-strike pattern or with modifications in the

8 Effects of Form-Focused Training on Running Biomechanics

stride length and cadence. Further studies in largersamples are needed to confirm these findings.

As can be seen in Figure 3, the changes in jointpowers in the form-focused group were in the directionof a decrease except for negative hip power. Furtherwork is needed to evaluate whether these changes aresignificant in larger cohorts and also to evaluatewhether these changes are independent of the changesin running speed. Specifically, we observed an increasein positive knee and ankle powers in the self-directedgroup after the training. The increase in positive kneepower (extensor during late stance) and positive anklepower (plantar flexor during late stance) could berelated to the increase in running speed in the self-directed group [22].

Goss et al [16] reported lower negative dorsiflexorwork in early stance, greater negative plantar flexorwork during early stance, and lower negative kneeextensor work in early stance in ChiRunners comparedwith rear-foot strikers. Similarly, the authors of anotherstudy reported greater negative plantar flexor work andlower negative knee extensor work in Pose (RomanovAcademy of Sports Science, Coral Gables, FL) runnerswith a forefoot strike pattern compared with runnerswith a rear-foot or mid-foot strike pattern [23]. In ourstudy, the form-focused group showed a decrease inpositive and knee and ankle joint powers, but the dif-ferences were not statistically significant. It may bepossible that these differences become greater with anincrease in experience.

There was a close to significant reduction in the kneeadduction moment after training in the ChiRunningtechniques. High knee adduction moment could leadto greater risk of patellofemoral joint pain and is knownto be related to greater loading over the medialcompartment. Hence, a potential reduction in kneeadduction moment could be important for individualswho have tibiofemoral or patellofemoral pain.

In summary, training in a running form that empha-sizes mid-foot strike, greater cadence, shorter stridelength, along with other characteristics could lead tochanges in running gait in untrained individuals. In thissmall sample, we observed potential improvements instrike index, and knee adduction moment that needfurther investigation in larger samples. Because this wasa pilot study to assess feasibility and gather preliminarydata for a larger trial, future studies are needed toassess the differences in various running form tech-niques, injury rates, and long-term compliance.

There are limitations to this work that need to beconsidered when the results are interpreted. The cohortincluded a small number of individuals, and, hence, theresults need to be confirmed in larger samples. Thesmall sample in this study could have limited our abilityto observe significant changes with the form-focusedtraining. We did not account for multiple comparisonswith the small sample size, considering the preliminary

nature of the work. Furthermore, the study includeduntrained individuals with increased blood pressure.Hence, the findings may not be generalized to activerunners.

Conclusions

Results from the pilot study demonstrate that thereare differences in the changes in running biomechanicsbetween individuals trained in the form-focusedtraining, which emphasizes mid-foot strike, greatercadence, and shorter stride length, compared withthose not trained in the running technique. These dif-ferences may be associated with reduced lower ex-tremity stress in individuals who are trained in theserunning techniques, but future studies are needed toconfirm these findings in larger samples.

Acknowledgments

The authors thank Cynthia Conti for assistance withmotion analysis data collections and Dr. Rebecca Fellinfor her thoughtful comments on the analyses of therunning biomechanics data. Funding support was pro-vided by the Mt. Zion Health Fund.

References

1. Running USA - State of the Sport Reports 2012-2013. 2012.2. Buist I, Bredeweg SW, Bessem B, van Mechelen W, Lemmink KA,

Diercks RL. Incidence and risk factors of running-related injuriesduring preparation for a 4-mile recreational running event. Br JSports Med 2010;44:598-604.

3. van Gent RN, Siem D, van Middelkoop M, van Os AG, Bierma-Zeinstra SM, Koes BW. Incidence and determinants of lower ex-tremity running injuries in long distance runners: A systematicreview. Br J Sports Med 2007;41:469-480; discussion 480.

4. Hreljac A, Marshall RN, Hume PA. Evaluation of lower extremityoveruse injury potential in runners. Med Sci Sports Exerc 2000;32:1635-1641.

5. Lopes AD, Hespanhol Junior LC, Yeung SS, Costa LO. What are themain running-related musculoskeletal injuries? A systematic re-view. Sports Med 2012;42:891-905.

6. Zadpoor AA, Nikooyan AA. The relationship between lower-extremity stress fractures and the ground reaction force: A sys-tematic review. Clin Biomech (Bristol, Avon) 2011;26:23-28.

7. Milner CE, Ferber R, Pollard CD, Hamill J, Davis IS. Biomechanicalfactors associated with tibial stress fracture in female runners.Med Sci Sports Exerc 2006;38:323-328.

8. Stefanyshyn DJ, Stergiou P, Lun VM, Meeuwisse WH, Worobets JT.Knee angular impulse as a predictor of patellofemoral pain inrunners. Am J Sports Med 2006;34:1844-1851.

9. Derrick TR, Hamill J, Caldwell GE. Energy absorption of impactsduring running at various stride lengths. Med Sci Sports Exerc 1998;30:128-135.

10. Edwards WB, Taylor D, Rudolphi TJ, Gillette JC, Derrick TR. Effectsof stride length and running mileage on a probabilistic stressfracture model. Med Sci Sports Exerc 2009;41:2177-2184.

11. Stergiou N, Bates BT, Kurz MJ. Subtalar and knee joint interactionduring running at various stride lengths. J Sports Med Phys Fitness2003;43:319-326.

9D. Kumar et al. / PM R XXX (2015) 1-9

12. Lieberman DE, Venkadesan M, Werbel WA, et al. Foot strike pat-terns and collision forces in habitually barefoot versus shod run-ners. Nature 2010;463:531-535.

13. Heiderscheit BC, Chumanov ES, Michalski MP, Wille CM, Ryan MB.Effects of step rate manipulation on joint mechanics duringrunning. Med Sci Sports Exerc 2011;43:296-302.

14. Goss DL, Gross MT. A review of mechanics and injury trends amongvarious running styles. US Army Med Dep J 2012:62-71.

15. Dreyer D. ChiRunning. New York, NY: Fireside; 2009.16. Goss DL, Gross MT. A comparison of negative joint work and ver-

tical ground reaction force loading rates between chi runnersand rearfoot striking runners. J Orthop Sports Phys Ther 2013;43:682-684.

17. Wang Y, Wang QJ. The prevalence of prehypertension and hyper-tension among US adults according to the new joint nationalcommittee guidelines: New challenges of the old problem. ArchIntern Med 2004;164:2126-2134.

18. Dickinson HO, Mason JM, Nicolson DJ, et al. Lifestyle interventionsto reduce raised blood pressure: A systematic review of random-ized controlled trials. J Hypertens 2006;24:215-233.

19. Thomas S, Reading J, Shephard RJ. Revision of the Physical ActivityReadiness Questionnaire (PAR-Q). Can J Sport Sci 1992;17:338-345.

20. Crowell HP, Davis IS. Gait retraining to reduce lower extremityloading in runners. Clin Biomech (Bristol, Avon) 2011;26:78-83.

21. Goss DL. A Comparison of Lower Extremity Joint Work and InitialLoading Rates Among Four Different Running Styles. Allied HealthSciences (Human Movement Science) [dissertion]. Chapel Hill, NC:University of North Carolina; 2013.

22. Schache AG, Blanch PD, Dorn TW, Brown NA, Rosemond D,Pandy MG. Effect of running speed on lower limb joint kinetics.Med Sci Sports Exerc 2011;43:1260-1271.

23. Arendse RE, Noakes TD, Azevedo LB, Romanov N, Schwellnus MP,Fletcher G. Reduced eccentric loading of the knee with the poserunning method. Med Sci Sports Exerc 2004;36:272-277.

Disclosure

D.K. Musculoskeletal Quantitative Imaging Research Group, Department ofRadiology and Biomedical Imaging, University of California, San Francisco, SanFrancisco, CA; and Division of Physical Therapy, College of Health Professions,Medical University of South Carolina, Charleston, SC. Address correspondenceto: D.K., 151B Rutledge Ave, MSC 962, Charleston, SC, 29425; e-mail:[email protected] related to this publication: grant, Mount Zion Health Fund, #20110449(money to institution)

K.M. Osher Center for Integrative Medicine, University of California San Fran-cisco, San Francisco, CADisclosure related to this publication: grant, Mount Zion Health Fund (money toinstitution); grant, NIH #T32AT003997 (money to Institution)

H.F. Osher Center for Integrative Medicine, University of California San Fran-cisco, San Francisco, CADisclosure related to this publication: grant, Mount Zion Health Fund (money toinstitution)

V.G. Osher Center for Integrative Medicine, University of California San Fran-cisco, San Francisco, CADisclosure related to this publication: grant, Mt. Zion Health Fund (money toinstitution)

A.L. Department of Orthopaedic Surgery, University of California San Francisco,San Francisco, CADisclosures outside this publication: board membership, SPORTZPEAK Inc;employment, SPORTZPEAK Inc; stock/stock options, SPORTZPEAK Inc.

R.B.S. Musculoskeletal Quantitative Imaging Research Group, Department ofRadiology and Biomedical Imaging and Department of Physical Therapy andRehabilitation Science, University of California San Francisco, San Francisco, CADisclosure related to this publication: grant, Mount Zion Health Fund #20110449(money to institution)

F.M.H. Osher Center for Integrative Medicine, University of California SanFrancisco, San Francisco, CADisclosure related to this publication: grant, NIH AT007827 (money to institution)

Funded by the Mt. Zion Health Fund.

Submitted for publication February 27, 2014; accepted January 21, 2015.