2015_Medial and Lateral Heel Whips Prevalence and Characteristics in Recreational runners

SAVE OFFLINE

PM R 7 (2015) 823-830

www.pmrjournal.org

Original Research

Medial and Lateral Heel Whips: Prevalence and Characteristics in Recreational Runners Richard B. Souza, PT, PhD, Nicolas Hatamiya, BA, Carly Martin, DPT, Andrew Aramaki, DPT, Brian Martinelli, DPT, Jamie Wong, DPT, Anthony Luke, MD, MPH

Abstract Objective: To investigate the prevalence and characteristics of recreational runners with medial and lateral heel whips. Design: Observational cohort study. Setting: Clinical research laboratory. Participants: A total of 256 recreationally active runners and joggers participated. Main Outcome Measures: High-definition video was acquired from a posterior view while runners ran at a self-selected pace on a treadmill. Heel whips, defined as the medial or lateral rotation of the foot in the transverse plane during initial swing, were measured with Dartfish software. Subjects were stratified by direction (medial and lateral) and severity (W_5-10 ¼ 5-10 degrees; W_10þ ¼ >10 degrees) of heel whip. Body mass index and gender comparisons, as well as measurement reliability, also were explored. Results: Mean heel whip angle across runners was 0.4 degrees (medial) with a standard deviation of 9.2 degrees. Of the 512 feet analyzed, 274 (54%) demonstrated a 5 degree whip or greater. There was a similar number of medial and lateral heel whips observed (27% each). Female runners were twice as likely to demonstrate a lateral heel whip of greater than 8.9 degrees. Overweight runners had more medially directed whips when compared with normal and underweight runners. Conclusions: More than half of the recreational runners studied were observed to have a medial or lateral heel whip of greater than 5 degrees. These data reveal the age, body mass index, and gender distribution of recreational runners with and without heel whips.

Introduction Long-distance running as a recreational sport has seen a substantial increase in popularity in recent years. According to the National Runner Survey (RunningUSA. org), in 2012 it was estimated that 1.85 million runners completed a half-marathon, representing a 15% increase from just a year previously. However, despite advances in running shoe technology and injury assessment capabilities, 30%-79% of runners are injured each year, without evidence of a marked decrease [1-3]. As popularity gains and injuries continue to occur, evaluation of these injuries and prospective injury prevention strategies are becoming increasingly important for the physical therapist and sports medicine clinician. A number of key variables have been identified and linked to common running-related injuries in

recent years. For example, hip kinematics have been implicated in injuries such as patellofemoral pain and iliotibial band syndrome [4-9]. In addition, stride impact forces and stride frequency also have been implicated in common running injuries [10-13]. Foot kinematics and shoe supports have been evaluated extensively in relation to running injuries [14-16]. However, one commonly observed running characteristic has received little attentiondmedial and lateral heel whips. The heel whip, defined here as a medial or lateral rotation of the foot in the transverse plane during initial swing, is named by the direction of the movement of the heel (ie, an external rotation of the foot during initial swing would be defined as a medial heel whip). Although there is some evidence of an appreciation for this variable in the prosthetic literature [17], there is little to no reports in the

1934-1482/$ - see front matter ª 2015 by the American Academy of Physical Medicine and Rehabilitation http://dx.doi.org/10.1016/j.pmrj.2015.02.016

824

Medial and Lateral Heel Whips

running literature. It is therefore the objective of this study to investigate the prevalence of medial and lateral heel whips in recreational runners, to describe the characteristics of individuals with medial and lateral heel whips, and to evaluate the reliability of measuring heel whips using standard 2-dimensional video analysis. Methods Subjects A total of 256 subjects were included in the study with bilateral analysis, resulting in 512 feet being analyzed (138 women and 118 men, mean # standard deviation [SD] height and weight ¼ 1.71 # 0.10 m and 67.2 # 11.1 kg, respectively). Subjects were recruited from the RunSafe program (SportzPeak, Inc) at the University of California, San Francisco. RunSafe is a sports wellness program for runners that focuses on personalized health advice, injury prevention, and performance enhancement. The program markets itself to running groups and the general public through local running clubs, running literature, and word-of-mouth. As such, the subjects for the current study consist of a wide range of demographics and performance levels. Consecutive runners attending the program between August 2010 and May 2011 were analyzed for the current investigation. Although injury status was not evaluated in the current investigation, it should be noted that many of the included runners had ongoing injuries at the time of data collection, consistent with common injuries seen in this population. All subjects signed an informed consent before participating in the study. The study was approved by the Committee on Human Research at the University of California, San Francisco. Instrumentation Participants wore their own running shoes for all testing procedures and ran at a self-selected speed on a standard treadmill (Model Biodex 600; Biodex Medical Systems, Inc., Shirley, NY) with no incline. Standard video was acquired from a high-definition video camera (HDRHC3 HDV 1080i Handycam Camcorder; Sony Electronics Inc., Laredo, TX) at 60 fields per second, positioned directly behind the treadmill. Camera location and angle relative to the treadmill were positioned identically for each acquisition to produce a standard video view for analysis. All video data was acquired and analyzed in Dartfish Video Software (Dartfish USA, Alpharetta, GA). Procedures Age, height, and weight for each subject were recorded before video acquisition. Next, each runner was positioned on the treadmill and told to “run at a

comfortable pace as though you are going for a long run.” Subjects were instructed to warm-up during a 2-minute period, after which they increased the speed of the treadmill to a comfortable self-selected speed. Once this speed was reached, and the comfort of the speed was confirmed with the runner, video acquisition began. A total of 60 seconds of video was recorded from a video camera position directly posterior to the treadmill at a fixed distance (11 feet) away from the back surface of the treadmill. The first 20 seconds of video included footage zoomed in from the knee joint down to the base of the treadmill, and was used for all analyses in the current investigation. Videos were stored on an external hard drive and analyzed with the use of Dartfish Video Software. Heel whips were quantified as a medial or lateral rotation of the foot in the transverse plane during initial swing. This was measured on the 2-dimensional video from the posterior view where the initial angle of the plantar surface of the shoe was measured relative to a vertical line just prior to toe-off in late stance (Figure 1, left). Next, the frame with the maximum amount of deviation in either the medial or lateral direction was identified, and again the angle between the plantar surface of the shoe relative to a vertical line was recorded (Figure 1, right). This second frame was required to occur within the first 170 milliseconds after toe-off. This requirement was selected as it was noted that several individuals demonstrated additional transverse plane motion of the foot in mid-swing as the knee continued to flex. However, the majority of subjects demonstrated peak angular deviation within the first 3-4 video frames after toe-off. The whip angle was calculated as the angle just before toe-off subtracted from the angle at maximum heel deviation. A positive number indicated a medial whip and negative number indicated a lateral whip. In instances when subjects demonstrated a combination of both medial and lateral whips, only the initial deviation was recorded (ie, if a subject exhibited a 5-degree lateral whip followed by a 7-degree medial whip, this was recorded as 5-degree lateral whip). This procedure was repeated on 3 consecutive strides for each subject. A subset of 29 subjects (58 data points) was measured a second time to perform a reliability assessment. During this second assessment, the tester remained blinded to the initial measurements for each subject. Data Analysis Medial and lateral whip measurements from 3 successive trials were averaged for all analyses. The mean and SD for the entire group was calculated. An arbitrary cut-off of #5 degrees was selected as criteria for normal heel deviation. Therefore, a runner with a medially or laterally directed whip greater than 5 degrees was

825

R.B. Souza et al. / PM R 7 (2015) 823-830

Figure 1. Examples of measurements made to calculate medial whip. The angle of the shoe sole relative to vertical for the initial frames (left) and the peak deviation (right) are recorded. The total angular excursion between initial frame and peak deviation frame is recorded as heel whip angle.

operationally defined as a runner with a medial or lateral heel whip. Severity of heel whip was further analyzed as follows: heel whip of 5-10 degrees ¼ W_5-10; greater than 10 degrees ¼ W_10þ. A secondary analysis was performed to determine the relationship between body mass index (BMI) and heel whips. Mean whip angles were calculated for underweight subjects (BMI < 20), for normal-weight subjects (BMI between 20 and 25), and for overweight subjects (BMI > 25). For the binary logistic generalized linear model, left and right heel whip angles were averaged, resulting in a single whip value for each runner. Statistical Analysis Descriptive statistics were used to describe the mean and variance of heel whips in this cohort of recreational runners. Prevalence of heel whip, stratified by severity, was described as percentage of cases observed. Additionally, characteristics of subjects with medial and lateral whips were reported as frequencies. Bonferroniadjusted independent-samples t-tests were performed to evaluate differences in characteristics between runners with W_10þ medial and lateral heel whips, as well as between subjects with whips and those without heel whips. Significance value will be adjusted to 0.0167 based on 3 planned comparisons: (1) W_10þ medial heel whip group vs W_10þ lateral heel whip group; (2) W_10þ medial heel whip group vs no heel whip group; and (3) W_10þ lateral heel whip group vs no heel whip group. Similarly, adjusted t tests were performed to evaluate differences across 3 BMI categories. Pearson’s correlations were used to explore relationships between BMI and average heel whip angle across all subjects. Gender odds ratio and 95% confidence intervals were calculated for W_10þ medial and lateral heel whips, and binary logistic generalized linear model was used to evaluate the likelihood of gender given an average

measured whip angle. Finally, intra-rater reliability of heel whip quantification was evaluated using intra-class correlation coefficients (ICC[3,3]), for runners across groups, as well as for runners with medial and lateral heel whips. All analyses were performed in SPSS software (SPSS Inc., Chicago, IL) with P value of .05. Results Prevalence Subject demographics are displayed in Table 1. The overall mean whip angle across all individuals was 0.4 degrees (medial) with a SD of 9.2 degrees (Figure 2). Of the 512 feet analyzed, 274 (54%) demonstrated a 5-degree whip or greater. There was a similar number of medial and lateral heel whips observed (medial: 136/512 ¼ 27%; lateral: 138/512 ¼ 27%). When stratified by whip severity, 17% (87/512) demonstrated a W_10þ (>10 degree) medial heel whip, and 10% (49/512) demonstrated a W_5-10 (5-10 degrees) medial heel whip. In contrast, 8% (40/512) of runners generated a W_10þ lateral heel whip, whereas 19% (98/512) generated a W_5-10 lateral heel whip. Of the 171 subjects that exhibited a heel whip on at least one side, the majority of cases involved bilateral presence of heel whips (103/171 ¼ 60%), with slightly fewer cases of unilateral whips observed (68/171 ¼ 40%). Only 2 cases (0.8%) were identified in which the runner had a medial Table 1 Subject characteristics Gender

N (%)

Age, y Mean # SD (range)

BMI, kg/m2, Mean # SD (range)

Male Female Total

118 (46) 138 (54) 256

42.3 # 10.9 (17-80) 41.9 # 9.7 (20-65) 42.1 # 10.3 (17-80)

23.6 # 2.3 (18.8-32.5) 22.2 # 2.9 (17.9-36.8) 22.9 # 2.7 (17.9-36.8)

BMI ¼ body mass index.

826

Medial and Lateral Heel Whips

10- degree medial heel whip were significantly older (44.2 vs 38.6 years) and had a greater BMI (24.0 vs 21.8 kg/m2) compared with subjects with a W_10þ lateral heel whip. In addition, there were a much larger percentage of females in the W_10þ lateral heel whip cohort (68% females, 32% males) compared with the W_10þ medial heel whip cohort (52% women, 48% men).

Figure 2. Frequency of heel whips in recreational runners. Negative values indicate lateral heel whips and positive values indicate medial whips.

whip on one side and a lateral whip on the other side, each greater than 5 degrees. Characteristics Intrasubject Variability Across the entire cohort, there was noted intrasubject variability both within a single limb and between left and right limbs. Within a single limb, the mean SD was 2.2 degrees across the 3 quantified strides. We observed a mean absolute difference angle between left and right feet of 4.2 degrees across all subjects. The maximum left-to-right difference observed within a single individual was 35.4 degrees (27.1 degree lateral whip on the left and 8.3 medial whip on the right). Medial Whips vs Lateral Whips Subjects with a W_5-10 medial heel whip were similar in age and BMI compared with subjects with a W_5-10 lateral heel whip (Table 2). In addition, there was a similar gender distribution between the groups of W_5-10 heel whips (42%-47% men, 53%-58% women). However, when comparing subjects in the W_10þ heel whip categories, we found that those with more than a

Medial and Lateral Whips vs No Whips Subjects with a W_5-10 medial heel whip were very similar in age, BMI, and gender distribution compared with those subjects with less than 5 degrees of whip (Table 2). Subjects with W_10þ medial heel whips, however, were similar in age (44.2 vs 42.0 years; P ¼ .104) but found to have a greater BMI (24.0 vs 22.7; P ¼ .0165) compared with subjects without heel whips. A similar gender distribution was observed in the W_10þ medial heel whip cohort compared with subjects without heel whips (48%-49% men, 51%-52% women). Subjects with a W_5-10 lateral heel whip were similar in age and BMI compared with subjects without whips (Table 2), but a larger percentage of women was noted in the cohort of subjects with W_5-10 lateral whips (58% vs 51% in the no-heel-whip group). Subjects with W_10þ lateral whips were noted to be insignificantly younger that those without heel whips (38.6 vs 42.0 years; P ¼ .06), and had significantly lower BMI (21.8 vs 22.7; P ¼ .003). In addition, a much larger percentage of women were noted in the group with W_10þ lateral heel whips compared with the group without heel whips (68% vs 51%). BMI Evaluation Runners who were underweight (BMI < 20) had a significantly lower heel whip angle, indicating a more lateral heel whip compared with overweight subjects (P < .001), with a similar trend observed compared with the normal BMI cohort (P ¼ .03; Table 3). Overweight runners had a significantly greater heel whip, indicating a more medially directed heel whip angle compared with both the underweight and normal-weight runners (P < .008). There were no differences in ages across BMI categories. Pearson’s correlation coefficient revealed a weak but significant relationship between BMI and average heel whip angle (r ¼ 0.25; P < .001; Figure 3),

Table 2 Descriptive statistics HW, severity

N (%)

Age, y Mean # SD

Gender, Male:Female(%)

BMI, kg/m2, Mean # SD

W_10þ medial HW (>10 degrees) W_5-10 medial HW (5-10 degrees) Minimal/ no HW (10 degrees)

87 49 238 98 40

44.2 41.2 42.0 42.3 38.6

42:45 23:26 117:121 41:57 13:27

24.0 23.0 22.7 22.4 21.8

(17) (10) (46) (19) (8)

# # # # #

11.1* 9.3 10.6 9.6 7.8

HW ¼ heel whip. * Significantly different than W_10þ lateral HW group. † Indicates significantly different than both W_10þ medial HW and W_10þ lateral HW groups.

(48:52) (47:53) (49:51) (42:58) (32:68)

# # # # #

3.4* 4.4 2.4† 2.5 2.3

827

R.B. Souza et al. / PM R 7 (2015) 823-830

Discussion

Table 3 BMI characteristics

BMI

N (%) 2

Heel Whip Angle, Degrees Age, y, Mean # SD Mean # SD

Underweight (25 kg/m2)

44.2 # 11.1 41.2 # 9.3 42.0 # 10.6

Heel whip angle is reported in degrees (medial is positive, negative is lateral). BMI ¼ body mass index. * Significantly different than overweight group. † Significantly different than both the normal-weight group and the underweight group.

with a greater BMI being associated with a medial heel whip. Gender Evaluation The odds of a female runner having a W_10þ lateral heel whip compared with a male runner were 1.86, with a 95% confidence interval between 0.94 and 3.69. The odds of a female runner having a W_10þ medial heel whip compared with a male were 0.90 with a 95% confidence interval between 0.57 and 1.43. A binary logistic generalized linear model was not significant for predicting gender based on heel whip angle (likelihood ratio c2 ¼ 3.34; P ¼ .188). The graphical prediction of gender when given whip angle is presented in Figure 4. Reliability With regard to measurement reproducibility, excellent reliability was observed across all measures (ICC ¼ 0.994; standard error of the mean ¼ 1.2 degrees; Figure 5). Furthermore, within measures of medial and lateral whips, excellent reliability also was observed (ICC ¼ 0.960 and 0.953, respectively, for medial and lateral whips).

This study revealed that more than half of all recreational runners included in the current investigation generated a medial or lateral heel whip of greater than 5 degrees. A similar frequency (w27%) of medial and lateral heel whips was observed in this cohort. However, there was over twice as many W_10þ medial heel whips (>10 degrees) compared with W_10þ lateral heel whips. Runners with W_10þ lateral heel whips were younger, had a lower BMI, and were more often female compared with those with W_10þ medial heel whips. A similar pattern was observed between subjects with W_10þ lateral heel whips and subjects without whips, although these failed to reach statistical significance (P ¼ .05-.06). Subjects with W_10þ medial heel whips had a greater BMI than subjects without heel whips but were similar in age and gender distribution. Finally, overweight runners had more medially directed whips compared with normal and underweight runners. These data reveal the age, BMI, and gender distribution of recreational runners with and without heel whips. It is not uncommon for clinicians skilled in observational running analysis, to recognize and make note of medial and lateral heel whips in runners. However, until now it was unknown how common heel whips are among recreational runners. In our study of 256 runners across a wide range of experience and age levels, more than half demonstrated a 5-degree heel whip in either the medial or lateral direction. As shown in Figure 2, the distribution of whips is roughly normal (skewness ¼ 0.384, kurtosis ¼ 0.698), with an average whip angle very close to neutral, and a similar number of runners with greater than 5 degrees of heel whip in each direction. The range of heel whips in this cohort was observed to be between 29 degrees lateral and 32 degrees medial. Although a majority of subjects demonstrating whips were observed to have bilateral whips, it was not uncommon to find unilateral heel

Likelihood of Being a Female Runner

Probability of Being Female

1

0.9

0.8

0.7

0.6

0.5

0.4 -30

-20

-10

0

10

20

30

Whip Angle (degrees)

Figure 3. Correlation between BMI and average heel whip angle (n ¼ 256); r ¼ 0.247; P < .001. BMI, body mass index.

Figure 4. Probability of being a female runner based upon heel whip angle.

828

Medial and Lateral Heel Whips

Repeatability of Heel Whip Measurements ICC(3,3) = .994; SEM=1.2 degrees

Measurement #2 (degrees)

40

R² = 0.977

20

-40

-20

0

0

20

40

-20

-40 Measurement #1 (degrees)

Figure 5. Scatterplot of heel whip data demonstrating excellent repeatability. ICC, intraclass correlation coefficient; SEM, standard error of the measurement.

whips as well. Of the 256 total subjects in the current study, only 33% (85/256) demonstrated less than a 5-degree heel whip on both feet. In fact, a bilateral heel whip of >5 degrees was observed in 40% of runners, and unilateral heel whips occurred in 27% of all runners studied. This highlights the varied presentation that a clinician might observe upon performing a running analysis. When evaluating the characteristics of subjects who demonstrated W_10þ medial heel whips, we found that these subjects had a greater BMI than either those with W_10þ lateral heel whips, or those without heel whips, although on average all groups had normal BMI values. Throughout the entire cohort of runners with W_10þ medial whips, the gender distribution was fairly even (48% men, 52% women), but data from our generalized linear model (Figure 4) suggest that this cohort may not be completely homogenous and subjects with the most medially directed heel whips may differ in gender distribution compared with those with heel whips closer to 10 degrees. It is not clear how greater BMI would directly affect medial whip angle. Future investigations should investigate this association to determine if this lies on a causal pathway, or is a simple association. Runners with W_10þ lateral heel whips were younger and had a lower BMI compared with runners with W_10þ medial heel whips. However, although statistical trends were noted (P ¼ .05-.06), these runners were not significantly different from runners without heel whips. In addition, there was a relatively high percentage of female runners who demonstrated W_10þ lateral heel whips. Gender differences in running mechanics have been evaluated in several previous studies. Ferber et al [18] reported that female recreational runners

demonstrate greater hip adduction, hip internal rotation, and knee abduction compared with male runners. In addition, various studies have reported that female runners demonstrate differences in neuromusclar activation patterns during running and jumping tasks in the lateral hamstrings, vastus lateralis, and peroneus longus compared with male runners [19,20]. However, in the current investigation, heel whip angle was similar between genders (males runners ¼ 1.0 degree medial whip vs female runners ¼ $0.1 degree lateral whip), suggesting that this kinematic pattern is not solely related to gender and may have additional underlying sources. Nonetheless, it is clear that subjects with the most W_10þ heel whips (both medial and lateral) tend to be female. A weak, yet statistically significant correlation was observed between BMI and heel whip angle (r ¼ 0.25; Figure 3). Overweight runners (BMI > 25) had a greater heel whip angle (indicating a medially directed heel whip) compared with both normal weight and underweight subjects (Table 3). In fact, although only 4 subjects in the current cohort had a BMI of greater than 30, the average heel whip angle was 13 degrees in this subgroup, with 6 of the 8 feet analyzed demonstrating a W_10þ medial heel whip. Because of the extremely low sample size of this cohort, we did not run statistics on these data, but this observation clearly needs to be explored in a larger study. In addition, the stability of this measurement needs to be investigated in runners whose BMI changes substantially over time. Binary logistic generalized linear model was used to evaluate the association of gender and heel whip angle. Figure 4 displays these findings with a clear indication that at both extremes of medial and lateral heel whips, the likelihood of being female greatly increases. Given the clear gender predisposition for common overuse injuries such as patellofemoral pain and iliotibial band syndrome [21,22], severely abnormal heel whips may be an indicator for faulty mechanics that can be used to identify at-risk runners. Obviously this assertion is speculative at this time and needs additional prospective investigations to either confirm or refute this suggestion. We observed excellent reliability with a 2-dimensional approach to measuring heel whips in recreational runners. On the basis of these data, one can be 95% confident that the measurement taken on one day is within 2.4 degrees of the measurement taken on a second day. Given the spread of data observed in the current investigation, this measurement should be able to accurately identify many of the subjects in the W_10þ heel whip categories. These findings are particularly translatable to the clinical therapy community because the methods employed consisted of standard videotaping of runners on a treadmill, and whip measurements were made with simple angle tools, available in many basic video analysis programs

R.B. Souza et al. / PM R 7 (2015) 823-830

that are available free of charge. After an initial training period, heel whip quantification is easily measured in 2-3 minutes per foot, which includes averaging measurements on 3 consecutive strides, allowing for clinical integration without significant time barriers. The current investigation reports on a new method for quantifying heel whips during running using standard videotape technology. To date, this is the first such study to describe the prevalence and characteristics of this pattern using any metrics in recreational runners. As such, the significance of these data remains unclear. The clinical significance of the heel whipdeither medial or lateraldis linked with its root cause, which at this time remains unknown. It is also unclear if this kinematic deviation is predictive of or associated with any known running injuries, as there has been little to no report of this in the running literature. Because we did not have injury data on these subjects, we are unable to explore this possibility. However, future investigations may consider exploring the effects of common running injuries that have been linked to frontal or transverse plane lower extremity kinematics or kinetics (such as patellofemoral pain or iliotibial band syndrome) on this variable. Often, and not surprisingly, investigations on running kinematics and kinetics have been limited to stance phase variables, because it is well established that the forces developed during this phase are much larger than swing phase forces. However, we speculate that although the observation described in the current manuscript is technically a swing phase variable, it may have a root cause in forces developed during stance phase. It is possible that this kinematic pattern is related to a kinetic variable that has been discussed in the running literaturedthe free vertical moment. This variable describes a torque about the vertical axis caused by friction between the foot and the ground during stance phase of running [23]. Several studies have investigated this variable in the running population. Holden and Cavanagh [23] reported that the vertical free moment was related to foot pronation in healthy runners. Furthermore, Milner et al [24] reported that the vertical free moment was significantly greater in runners with a history of tibial stress fractures, compared with healthy runners. In fact, their analysis showed that the peak free moment was able to successfully predict a history of tibial stress fractures in 66% of cases. Their data reveal variations among runners into both adduction-free moment and abduction-free moments, the result of which would likely result in medial and lateral whips, respectively. These peaks were larger in the subjects with a history of tibial stress fractures. However, it should be noted that the average free moment across the 2 study groups peaked at mid-stance and approached zero at toe-off. Therefore, the resultant rotational forces

829

imparted to the lower extremity upon unweighting, would likely be fairly minimal compared with the forces present during stance. It is possible that the subjects observed in the current investigation that demonstrated large heel whips continued to store potential energy upon initial swing, which was released in the form of a heel whip. Future studies should evaluate the relationship between the vertical free moment and heel whips in runners with and without lower-extremity pathology. There are several limitations to the current investigation that need to be noted. First, this is a fairly small study sample to characterize running qualities in a diverse population. Therefore, care should be taken in extrapolating these findings to other populations. In addition, the recruitment center was a pay-for-service wellness facility, and as such, the cohort consisted of runners motivated to seek consultation on their running form, and with enough financial means to afford the service. It is unknown whether these factors influenced the results of the current observation, but it is certainly possible. Next, footwear was not provided, and therefore the cohort consisted of a variety of shoe brands and levels of support. It is currently unknown whether footwear would have an impact of heel whip magnitude. Finally, the method of kinematic quantification for the current investigation involved transverse plane angular measurements recorded on two-dimensional videotape. One potential source of error comes from measuring an angle of different objects from different distances (parallax error). Although the distance from the treadmill to the camera remained the same for all cases, the exact location of the runner on the treadmill belt was not controlled. Therefore, there is some variability in the distance between the camera and the foot of each runner, resulting is a small error in the whip angle calculation. However, our reliability data revealed excellent repeatability, providing some support for the measurement, although the validity of this measurement in 3 dimensions remains unknown. Conclusion More than half of the recreational runners studied were observed to have a medial or lateral heel whip of greater than 5 degrees. There was over twice as many W_10þ medial heel whips as W_10þ lateral heel whips. The cohort with W_10þ lateral heel whips was younger and had a lower BMI compared with those with W_10þ medial heel whips. Female runners were twice as likely to demonstrate a lateral heel whip of greater than 8.9 degrees compared with male runners. Finally, overweight runners had more medially directed whips compared with normal and underweight runners. These data reveal the age, BMI, and gender distribution of recreational runners with and without heel whips.

830

Medial and Lateral Heel Whips

Acknowledgments We would like to thank Cameron Sprowles and Cynthia Conti for their efforts in data acquisition on this project. References 1. Johnston CA, Taunton JE, Lloyd-Smith DR, McKenzie DC. Preventing running injuries. Practical approach for family doctors. Can Fam Physician 2003;49:1101-1109. 2. Bovens AM, Janssen GM, Vermeer HG, Hoeberigs JH, Janssen MP, Verstappen FT. Occurrence of running injuries in adults following a supervised training program. Int J Sports Med 1989;10(Suppl 3): S186-S190. 3. Macera CA, Pate RR, Powell KE, Jackson KL, Kendrick JS, Craven TE. Predicting lower-extremity injuries among habitual runners. Arch Intern Med 1989;149:2565-2568. 4. Souza R, Powers CM. Differences in hip kinematics, muscle strength, and muscle activation between subjects with and without patellofemoral pain. J Orthop Sports Phys Ther 2009;39:12-19. 5. Souza RB, Draper CE, Fredericson M, Powers CM. Femur rotation and patellofemoral joint kinematics: A weight-bearing magnetic resonance imaging analysis. J Orthop Sports Phys Ther 2010;40: 277-285. 6. Souza RB, Powers CM. Predictors of hip internal rotation during running: An evaluation of hip strength and femoral structure in women with and without patellofemoral pain. Am J Sports Med 2009;37:579-587. 7. Noehren B, Hamill J, Davis I. Prospective evidence for a hip etiology in patellofemoral pain. Med Sci Sports Exerc 2013;45: 1120-1124. 8. Willson JD, Davis IS. Lower extremity mechanics of females with and without patellofemoral pain across activities with progressively greater task demands. Clin Biomech (Bristol, Avon) 2008;23: 203-211. 9. Willy RW, Manal KT, Witvrouw EE, Davis IS. Are mechanics different between male and female runners with patellofemoral pain? Med Sci Sports Exerc 2012;44:2165-2171. 10. Heiderscheit BC, Chumanov ES, Michalski MP, Wille CM, Ryan MB. Effects of step rate manipulation on joint mechanics during running. Med Sci Sports Exerc 2011;43:296-302.

11. Hreljac A, Marshall RN, Hume PA. Evaluation of lower extremity overuse injury potential in runners. Med Sci Sports Exerc 2000;32: 1635-1641. 12. Milner CE, Hamill J, Davis I. Are knee mechanics during early stance related to tibial stress fracture in runners? Clin Biomech (Bristol, Avon) 2007;22:697-703. 13. Edwards WB, Taylor D, Rudolphi TJ, Gillette JC, Derrick TR. Effects of stride length and running mileage on a probabilistic stress fracture model. Med Sci Sports Exerc 2009;41:2177-2184. 14. Rodrigues P, Chang R, Tenbroek T, Hamill J. Medially posted insoles consistently influence foot pronation in runners with and without anterior knee pain. Gait Posture 2013;37:526-531. 15. Zifchock RA, Davis I. A comparison of semi-custom and custom foot orthotic devices in high- and low-arched individuals during walking. Clin Biomech (Bristol, Avon) 2008;23:1287-1293. 16. Vicenzino B, Griffiths SR, Griffiths LA, Hadley A. Effect of antipronation tape and temporary orthotic on vertical navicular height before and after exercise. J Orthop Sports Phys Ther 2000; 30:333-339. 17. Menard MR, McBride ME, Sanderson DJ, Murray DD. Comparative biomechanical analysis of energy-storing prosthetic feet. Arch Phys Med Rehabil 1992;73:451-458. 18. Ferber R, Davis IM, Williams DS 3rd. Gender differences in lower extremity mechanics during running. Clin Biomech (Bristol, Avon) 2003;18:350-357. 19. Palmieri-Smith RM, McLean SG, Ashton-Miller JA, Wojtys EM. Association of quadriceps and hamstrings cocontraction patterns with knee joint loading. J Athl Train 2009;44:256-263. 20. Baur H, Hirschmuller A, Cassel M, Muller S, Mayer F. Genderspecific neuromuscular activity of the M. peroneus longus in healthy runnersda descriptive laboratory study. Clin Biomech (Bristol, Avon) 2010;25:938-943. 21. Taunton JE, Ryan MB, Clement DB, McKenzie DC, Lloyd-Smith DR, Zumbo BD. A retrospective case-control analysis of 2002 running injuries. Br J Sports Med 2002;36:95-101. 22. Tenforde AS, Sayres LC, McCurdy ML, Collado H, Sainani KL, Fredericson M. Overuse injuries in high school runners: Lifetime prevalence and prevention strategies. PM R 2011;3:125-131. quiz 131. 23. Holden JP, Cavanagh PR. The free moment of ground reaction in distance running and its changes with pronation. J Biomech 1991; 24:887-897. 24. Milner CE, Davis IS, Hamill J. Free moment as a predictor of tibial stress fracture in distance runners. J Biomech 2006;39:2819-2825.

Disclosure R.B.S. Department of Physical Therapy and Rehabilitation Science, University of California, San Francisco, 185 Berry Street, Suite 350, San Francisco, CA 94107; and Department of Orthopaedic Surgery, University of California, San Francisco, San Francisco, CA. Address correspondence to: R.B.S.; e-mail: richard.souza@ ucsf.edu Disclosure: nothing to disclose N.H. Department of Orthopaedic Surgery, University of California, San Francisco, San Francisco, CA Disclosure: nothing to disclose C.M. Department of Physical Therapy and Rehabilitation Science, University of California, San Francisco, San Francisco, CA Disclosure: nothing to disclose A.A. Department of Physical Therapy and Rehabilitation Science, University of California, San Francisco, San Francisco, CA Disclosure: nothing to disclose

B.M. Department of Physical Therapy and Rehabilitation Science, University of California, San Francisco, San Francisco, CA Disclosure: nothing to disclose J.W. Department of Physical Therapy and Rehabilitation Science, University of California, San Francisco, San Francisco, CA Disclosure: nothing to disclose A.L. Department of Orthopaedic Surgery, University of California, San Francisco, San Francisco, CA Disclosure outside this publication: stock/stock options, founder, RunSafe, SportZPeak, Inc Submitted for publication March 24, 2014; accepted February 28, 2015.