- Email: [email protected]
Sagittal and frontal plane joint mechanics throughout the stance phase of walking in adolescents who are obese
Gait & Posture 32 (2010) 263–268
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Sagittal and frontal plane joint mechanics throughout the stance phase of walking in adolescents who are obese A.G. McMillan a,*, A.M.E. Pulver a, D.N. Collier b, D.S.B. Williams a a b
Department of Physical Therapy, 2410E Health Sciences Building, East Carolina University, Greenville, NC 27834, United States Department of Pediatrics, Brody School of Medicine at East Carolina University, Greenville, NC, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 September 2009 Received in revised form 25 March 2010 Accepted 8 May 2010
The incidence of obesity has increased dramatically in children and adolescents, and with this comes health risks typically associated with adult obesity. Among those health consequences are musculoskeletal damage and pain. Previous studies have demonstrated inconsistent effects of increased body mass on movement patterns in adults and children who are obese. The purpose of this study was to investigate frontal and sagittal plane mechanics during walking in adolescents who were obese. Adolescents (12–17 years) who were obese were recruited from a weight management program, and healthy weight peers (matched for age, race and gender) were recruited from the community. Three-dimensional motion analysis of the lower extremities was performed during walking. Analysis of kinematic and kinetic data from 36 adolescents who were obese and healthy weight revealed significant differences in mechanics at all lower extremity joints in both sagittal and frontal planes. Subjects who were obese seemed to use movement strategies that minimized joint moments, especially at the hip and knee during walking. The lower extremity mechanics during walking in the subjects who were obese raise concerns about maintenance of structural integrity of the lower extremity joints over time, given the repeated high stresses across the joints even with walking. Neither the long term consequences of these atypical movement patterns, nor the ability to alter these patterns through therapeutic activities or weight loss has been investigated in adolescents who are obese. ß 2010 Elsevier B.V. All rights reserved.
Keywords: Obesity Adolescents Walking Biomechanics
1. Introduction Childhood obesity, defined by the Centers for Disease Control and Prevention as having a body mass index (BMI) greater than the 95th percentile for age and gender, has become an epidemic, with 17.4% of all adolescents (12–19 years) in the United States classified as obese [1]. Chronic conditions such as cardiovascular disease and diabetes, commonly seen in adults who are obese, are now seen in adolescents who are obese [2,3]. Musculoskeletal conditions, including osteoarthritis, low back pain, and soft tissue injury, are often associated with obesity [4,5]. Increased body mass, with increased forces across weight-bearing joints, has been causally implicated in many of these musculoskeletal conditions [5]. Forces on joint surfaces are increased during any weightbearing activity, including walking. Increased body mass may increase risk of damage and injury to joint surfaces and other musculoskeletal structures with repetitive loading during weightbearing activities. Many have hypothesized that movement patterns are significantly different in individuals who are obese.
* Corresponding author. Tel.: +1 252 744 6232; fax: +1 252 744 6240. E-mail address: [email protected] (A.G. McMillan). 0966-6362/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2010.05.008
While most studies report that children and adults who are obese walk slower, with a wider step width, shorter steps, and increased double support time/decreased single limb support time [6–9]. Nantel et al. [10] reported no difference in gait velocity, cadence, stride length or double limb support time in obese versus healthy weight (HW) children. Reports of kinematic and kinetic characteristics of walking in individuals who are obese are inconsistent. Peak knee flexion angles during stance phase of walking have been reported as being lower in children and adults who are obese [11,12] while others reported no difference in these angles [8,9]. Peak plantarflexion angles have been reported as lower [9] and higher [10] in obese versus HW subjects. The lower plantarflexion angles were noted in subjects with significantly slower gait velocity [9]. No differences in sagittal plane hip and knee moments have been found between obese and HW subjects during walking [8,11,12]. Compared to HW adults, sagittal plane ankle moments of obese adults have been reported as lower [8] and higher [12] during walking. Nantel et al. [10] reported the only difference in lower extremity sagittal plane kinetics during walking was an earlier shift from hip extension moment to hip flexion moment in obese versus HW children. Less data is available for frontal plane biomechanics during walking. Compared to HW individuals, peak hip frontal plane
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
264
angles during stance phase have been reported as significantly more abducted in adults who are obese [9], and as significantly more adducted in children who are obese [13]. The greater hip abduction angles were reported in subjects who walked significantly slower than their lean peers [9]. Greater knee valgus and rearfoot inversion during walking has been reported in children who are obese (compared to HW peers) [13], while studies have alternately reported higher knee adduction moments [13] and knee abduction moments [11] in obese versus HW children. In summary, a variety of investigations of the effects of obesity on characteristics of walking have yielded inconsistent results. Studies to date have included adults and children (8–13 years old). Frontal and sagittal plane biomechanics in hip, knee and ankle at specific points during the stance phase of walking in adolescents who are obese have not been reported. The purpose of this study was to examine the sagittal and frontal plane lower extremity biomechanics during walking in adolescents who were obese versus HW peers. 2. Methods 2.1. Subjects Both male and female adolescents who were obese and HW were recruited for this study. Inclusion criteria were: age (12–17 years), BMI (less than the 85th percentile for age and gender for HW group; and greater than 95th percentile for age and gender for obese group). Potential subjects were excluded if they had musculoskeletal, neuromuscular, and/or cardiopulmonary conditions other than obesity that would limit movement or compromise safety. Participants who were obese were recruited from a local healthy weight program. Healthy weight participants were recruited from the community and matched by age, gender, and race. Written informed consent by a parent/legal guardian and written assent by the adolescent were obtained for each participant prior to data collection. This study was approved by the University and Medical Center Institutional Review Board. 2.2. Equipment Kinematic data were collected using eight Hawk infrared digital cameras (Motion Analysis Corp., Santa Rosa, CA) with a sample rate of 120 Hz. Force data were collected using two synchronized force plates (AMTI, Watertown, MA) with a sampling rate of 960 Hz. Retro-reflective spherical markers were attached directly to the skin on each subject’s trunk and lower extremities at the sternum, T1, T10, L5/ S1, iliac crests, acromion processes, greater trochanters, medial and lateral femoral condyles, medial and lateral malleoli, posterior and lateral heels, and 1st and 5th metatarsal heads. Rigid arrays of markers were also placed on thighs and lower legs. Anatomical/joint markers (those defining joint centers and segment coordinate axes) were left on only during static calibration trial; 22 tracking markers remained on the subject’s calcaneus, shank, thigh, pelvis, and trunk during walking trials [13]. 2.3. Experimental procedures Participants were instructed to walk down a 15-foot walkway at a self-selected pace without looking down at the ground. In order to account for any differences in lower extremity mechanics that may exist due to differences in gait velocity, subjects were further matched on gait velocity. This resulted in 18 subjects in each group (Table 1). Each participant performed warm-up trials prior to data collection to establish a starting point so that each foot hit one of the force platforms during walking trials. Data collection continued until the subject completed at least 10 successful trials in which each foot contacted a force platform with no apparent change in gait pattern to purposefully hit the platforms.
Table 2 Sagittal plane kinematic and kinetic variables, with mean (SD) and p values. HW: healthy weight. Obese Sagittal plane kinematic variables (8) Ankle angle at initial 0.87 contact Ankle peak angle during 7.77 early stance Ankle peak angle during 7.19 late stance Ankle angle at toe off 9.99 Knee angle at initial contact 1.38 Knee peak angle during 11.26 early stance Knee peak angle during 4.17 midstance Knee angle at toe off 40.82 Hip angle at initial contact 18.01 Hip peak angle during 12.45 late stance Hip angle at toe off 7.76
HW
p-Value
(6.01)
0.76 (2.77)
0.15
(5.32)
4.69 (2.21)
0.02
(5.17)
7.18 (3.31)
0.50
(5.08) (7.35) (7.03)
6.82 (4.09) 7.10 (3.41) 16.09 (6.35)
(7.93)
7.22 (4.72)
0.02 0.003 0.02 0.09
(7.90) (10.50) (11.04)
37.28 (5.28) 30.47 (9.62) 4.69 (9.90)
(11.73)
1.00 (9.61)
0.01
0.11 (0.04)
0.44
Sagittal plane kinetic variables (Nm/kg m) Ankle peak moment during 0.11 (0.05) early stance Ankle peak moment during 0.67 (0.13) late stance Knee peak moment at initial 0.19 (0.06) contact Knee peak moment during 0.20 (0.14) early stance Knee peak moment during 0.10 (0.14) late stance Hip peak moment at 0.43 (0.12) initial contact Hip peak moment during 0.37 (0.16) late stance
0.06 0.000 0.02
0.88 (0.07)
0.000
0.28 (0.11)
0.001
0.12 (0.14)
0.05
0.31 (0.11)
0.000
0.72 (0.23)
0.000
0.24 (0.08)
0.002
gait (0% = initial contact, 100% = toe off). For subjects who were obese, hip joint centers were defined using a modified model to account for excessive subcutaneous tissue that would have significantly offset the location of the hip joint centers in these subjects. Specifically, the hip joint centers were calculated by first placing the anatomical markers on the skin over the greater trochanters. Anthropometric measures were then taken with calipers to approximate how much soft tissue was over the greater trochanters. This number was then used as an offset number and it was entered into the model file for Visual 3D. Virtual greater trochanter markers were then established and used to determine hip joint centers. Finally, hip joint centers were estimated by taking 25% of distance from each virtual greater trochanter marker. Moments were normalized to subject’s mass and height. Sagittal and frontal plane angles and moments at specific events (initial contact, toe off), and peak angles and moments during early stance (1st 30%), midstance (40–60%), and late stance (last 30%) were selected for statistical analysis. Data were averaged across all trials for each subject, and group averages were calculated and used for statistical analyses. Student’s t-tests were used to determine the differences between the obese and HW groups on the variables of interest. An a value of 0.05 was selected for statistical significance. Bonferroni corrections were performed due to multiple comparisons, resulting in corrected p value of 0.005 for kinematic data and 0.007 for kinetic data.
2.4. Data analysis
3. Results
Raw coordinate data were smoothed using a second order recursive Butterworth filter at 12 Hz for kinematics and 50 Hz for kinetics. EvaRT v5.0.4. Software (Motion Analysis Corp., Santa Rosa, CA) was used for data collection and initial processing of the raw data. Further processing was performed using Visual 3D software (C-Motion, Inc., Germantown, MD). Trials were normalized to the stance phase of
Sagittal plane. Subjects who were obese had a significantly lower plantarflexion moment during late stance compared to the HW group (Table 2 and Fig. 1D). The obese group had significantly less knee flexion at initial contact (Table 2 and Fig. 1B). In general
Table 1 Subjects characteristics.
Obese group (n = 18) HW group (n = 18) p-Value
Female/Male
African-American/ Caucasian
Age (years)
Height (m)
Weight (kg)
Actual BMI
BMI % (for age/gender)
BMI z-score
17/1 13/5
15/3 14/4
15.0 (1.5) 14.6 (1.8) 0.23
1.6 (0.1) 1.6 (0.1) 0.10
121.2 (30.8) 53.2 (7.0) 0.000
44.6 (10.2) 20.3 (2.0) 0.000
> 99 55.1 (22.8)
2.54 (0.34) 0.15 (0.65) 0.000
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
265
Fig. 1. Sagittal plane lower extremity kinematics (A–C) and kinetics (D–F) during stance phase of walking in adolescents who are obese (black) and healthy weight (HW, gray). Positive values are ankle dorsiflexion, knee extension, hip flexion. Solid lines = mean, dotted lines = 1SD.
the subjects who were obese maintained a relatively more extended knee throughout stance. Knee flexion moment was significantly lower in the obese group at initial contact and in late stance (Table 2 and Fig. 1E). At the hip, subjects who were obese remained less flexed/more extended than the HW group throughout stance, but this difference was only significant at initial contact (Table 2 and Fig. 1C). Subjects who were obese had significantly lower hip extension moments at initial contact, significantly higher hip flexion moments during late stance, and exhibited an earlier transition to hip flexion moment, compared to the HW group (Table 2 and Fig. 1F). Frontal plane. Subjects who were obese had a significantly lower rearfoot inversion moment during late stance, compared to their HW peers (Table 3 and Fig. 2D). At the knee, HW subjects had a neutral to adducted knee position, while subjects who were obese maintained an abducted (valgus) knee position throughout stance (Table 3 and Fig. 2B). Peak frontal plane knee angles were significantly different between groups at initial contact, during
early stance, and during late stance (Table 3 and Fig. 2B). Subjects who were obese had significantly greater frontal plane excursion between peak knee angle during early stance and peak angle during late stance compared to their HW peers (Table 3 and Fig. 2B). Frontal plane knee moments were also significantly different between groups, with subjects who were obese exhibiting lower knee abduction moments throughout much of stance (significant during all phases of stance; Table 3 and Fig. 2E). Hip angles were not significantly different between groups, with both groups exhibiting adduction then abduction during stance (Table 3 and Fig. 2C). The obese group did exhibit a significantly lower hip abduction moment during early stance (Table 3 and Fig. 2F). 4. Discussion The purpose of this study was to investigate the sagittal and frontal plane lower extremity biomechanics during walking in adolescents who were obese versus adolescents who were HW.
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
266
Table 3 Frontal plane kinematic and kinetic variables, with mean (SD) and p values. HW: healthy weight. Obese Sagittal plane kinematic variables (8) Rearfoot angle at 10.57 (3.92) initial contact Rearfoot angle peak 1.96 (3.47) midstance Rearfoot excursion 9.88 (2.69) midstance ! late stance Knee angle at 4.12 (4.37) initial contact Knee angle peak 0.76 (4.32) during early stance Knee excursion initial 3.36 (1.70) contact ! early stance Knee angle peak during 9.55 (7.62) late stance Knee excursion 8.79 (4.49) loading ! late stance Hip angle peak during 7.00 (3.47) early stance Hip angle excursion initial 6.12 (2.85) contact ! early stance Hip angle peak during 7.06 (3.80) late stance Frontal plane kinetic variables (Nm/kg m) Rearfoot moment amplitude 0.03 during early stance 0.07 Rearfoot moment peak amplitude during late stance Knee moment peak amplitude 0.03 during early stance Knee moment peak amplitude 0.16 during midstance Knee moment peak amplitude 0.14 during late stance Hip moment peak amplitude 0.42 during early stance Hip moment peak amplitude 0.47 during late stance
HW
p-Value
8.26 (4.88)
0.13
0.27 (5.03)
0.25
10.73 (3.34)
0.49
0.35 (2.46)
0.003
2.99 (2.73)
0.004
3.35 (1.35)
0.98
0.80 (3.94)
0.000
3.79 (2.90)
0.000
6.72 (4.20)
0.89
6.57 (2.95)
0.65
5.44 (3.18)
0.18
(0.01)
0.02 (0.02)
0.77
(0.03)
0.11 (0.02)
0.000
(0.03)
0.07(0.06)
0.003
(0.06)
0.30 (0.09)
0.000
(0.06)
0.27 (0.09)
0.000
(0.12)
0.55 (0.13)
0.003
(0.19)
0.54 (0.09)
0.18
Significant differences between groups were found in sagittal and frontal plane kinematics and kinetics at all three lower extremity joints during walking. Sagittal plane. The only significant difference between groups at the ankle in the sagittal plane was the peak plantarflexion moment in late stance, which was lower in the obese group. Other reports of sagittal plane ankle kinematics and kinetics in obese (versus HW) subjects have been inconsistent: no differences in motion, but lower muscle moments at the ankle [8]; lower plantarflexion angles and greater dorsiflexion angles [9]; greater plantarflexion motion and higher ankle moments [12]; lower plantarflexion moments [11]. Lower plantarflexor moments seen in this and other studies might be due to relative weakness, as the plantarflexors are partial contributors to ankle joint moments in the sagittal plane. Lower plantarflexion moment in late stance results in decreased push-off which will likely require increased contributions at other joints. The increased hip flexion moment seen in late stance in the obese subjects in this study is likely a compensation to pull the limb into swing rather than push it through with the plantarflexors. Several significant findings were noted in the sagittal plane at the knee joint. In general, subjects who were obese had less flexion during early stance and lower flexion moments at the knee at initial contact and during late stance. The knee angle findings are similar to those previously reported [11,12], though others reported no significant differences in knee sagittal plane angles [8,9]. Subjects who are obese may keep the knee less flexed to
compensate for potential knee extensor weakness, or less effective knee extensor pull given their more abducted knee alignment in stance. A relatively extended knee during stance may also be a compensation for instability within the knee joint structure. Neither knee extensor strength nor knee stability were directly measured in this study. Subjects who were obese had significantly less hip flexion and lower hip extension moments at initial contact, and higher hip flexion moment during late stance, compared to the HW group. Others have reported no difference in sagittal plane kinematics and kinetics at the hip [8,9,12]. Obese subjects in the current study also moved much earlier (during early stance) into hip extension with a concurrent hip flexion moment. This could be a compensation for relatively weak hip extensors: extension at the hip during stance brings the line of force closer to or behind the hip joint, creating an external extensor moment at the hip and thus requiring subjects to use less hip extensor strength during stance. Hip extensor strength was not reported in this study. Frontal plane. Unlike previous findings [13], rearfoot kinematics in the frontal plane were not different between groups, perhaps due to inclusion of females in the current study. Further investigation of differences in mechanics between males and females who are obese is warranted. Subjects who were obese did exhibit significantly lower rearfoot inversion moment in late stance, which may be due to relative weakness in the muscles controlling rearfoot eversion. While rearfoot motion was not different between groups, even a slight change in the lateral position of the ground reaction force relative to the joint center would likely explain changes in joint moments as reported here. As reported previously [13], subjects who were obese maintained an abducted (valgus) position at the knee throughout stance, while HW subjects remained neutral/adducted at the knee. Subjects who were obese also had significantly greater frontal plane knee excursion during stance. While in a previous study boys who were obese exhibited knee adduction moments during early and late stance and knee abduction moments through midstance, in the current study obese subjects’ knee moments remained in abduction throughout stance, and were significantly lower than those of HW subjects. Subjects in Gushue et al. [11] also exhibited knee abduction moments, though these moments were greater in subjects who were obese (versus HW). Differences in the current results may be partially due to the model used to calculate hip joint centers being based on actual anthropometric measurements. The repetitive stresses on the knee joint structures related to the greater frontal plane excursion and moments during stance is concerning due to the potential for damage to knee joint structures, pain, limited motion, and resultant disability. Frontal plane hip angles were not different between groups, and only the hip abduction moment during early stance was significantly different between groups. Differences from previous data [13] may be partially due to calculation of hip joint centers. Spyropoulos et al. [9] reported that adults who were obese maintained hip abduction during stance, but those data were based on two-dimensional skin markers (thus not taking into account the amount of subcutaneous tissue present in obese subjects). The lower hip abduction moment during early stance suggests relative hip abductor weakness, though data is not available at this time to confirm this suggestion. 5. Limitations Both males and females were recruited for this study in order to improve the generalizability of the results. While combining male and female subjects does introduce a source of potential variability, there were too few males actually enrolled in the
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
267
Fig. 2. Frontal plane lower extremity kinematics (A–C) and kinetics (D–F) during stance phase of walking in adolescents who are obese (black lines) and healthy weight (HW, gray lines). Positive values are rearfoot eversion, knee adduction, hip adduction. Solid lines = mean, dotted lines = 1SD.
current study to compare them separately. Potential differences in movement characteristics between males and females who are obese will need further study. Skin movement and marker placement errors are potential confounders of movement data, especially in subjects who are obese. Several methods were employed in the current study in an attempt to minimize these errors. The same investigator placed all markers on all participants to account for differences in marker placement due to investigators or days [14]. To limit skin and tissue motion, markers were attached to rigid shells, which were in turn attached to tight circumferential wraps on the distal portion of the tibia and femur. This method and location of attachment has been shown to minimize soft tissue motion relative to bone [15]. The choice was made not to use tight fitting shorts or other material around the hip to try to decrease this marker movement because the binding or stabilizing effect of the material itself could lead to alterations in gait characteristics. We are confident that the differences observed between groups were minimally due to skin movement or to marker placement.
6. Summary Subjects who were obese exhibited significant differences in both sagittal and frontal kinematics and kinetics of the lower extremities during the stance phase of walking. This was especially true at the knee, a joint which is prone to damage and injury. We hypothesize that muscle weakness is one potential cause of these movement differences, and are currently investigating the correlations between muscle strength at the hip and knee and the observed gait deviations in this group of subjects. An additional investigation is focusing on resistive strength training of lower extremity muscles implicated in the observed gait pattern. Because these movement differences are noted even in a relatively low stress movement such as walking, and because adolescents who are obese must walk on a daily basis (and in fact are often encouraged to walk to increase their physical activity), it is important to determine the true mechanisms behind this atypical gait pattern, and how/if these mechanisms can be addressed clinically. The long term consequences of atypical gait patterns in obese adolescents have yet to be identified.
268
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
Acknowledgements This study was funded in part by a Research Development Award to Dr. Gross McMillan from East Carolina University. These funds were used for supplies, subject reimbursement, and research assistant salary. Conflict of interest statement None of the authors on this manuscript had or have any financial and personal relationships with other people or organizations that could inappropriately influence (bias) this work. References [1] Centers for Disease Control and Prevention. Overweight and obesity: childhood overweight. http://www.cdc.gov/nccdphp/dnpa/obesity/childhood/ index.htm. Accessed August 24, 2008. [2] Patterson RE, Frank LL, Kristal AR, White E. A comprehensive examination of health conditions associated with obesity in older adults. Am J Prev Med 2004;27:385–90. [3] Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord 1993;23:S2–11.
[4] Kortt M, Baldry J. The association between musculoskeletal disorders and obesity. Aust Health Rev 2002;25:207–14. [5] Anandacoomarasamy A, Caterson I, Sambrook P, Fransen M, March L. The impact of obesity on the musculoskeletal system. Int J Obes 2008;32:211–22. [6] Hills AP, Parker AW. Gait characteristics of obese children. Arch Phys Med Rehabil 1991;72:403–7. [7] Hills AP, Parker AW. Locomotor characteristics of obese children. Child Care Health Dev 1992;18:9–34. [8] Browning RC, Kram R. Effects of obesity on the biomechanics of walking at different speeds. Med Sci Sports Exerc 2007;39:1632–41. [9] Spyropoulos P, Pisciotta JC, Pavlou KN, Cairns MA, Simon SR. Biomechanical gait analysis in obese men. Arch Phys Med Rehabil 1991;72:1065–70. [10] Nantel J, Brochu M, Prince F. Locomotor strategies in obese and non-obese children. Obesity 2006;14:1789–94. [11] Gushue DL, Houck J, Lerner AL. Effects of childhood obesity on three-dimensional knee joint biomechanics during walking. J Pediatr Orthop 2005;25: 763–8. [12] DeVita P, Hortobagyi T. Obesity is not associated with increased knee joint torque and power during level walking. J Biomech 2003;36:1355–62. [13] McMillan AG, Auman NL, Collier DN, Williams DSB. Frontal plane lower extremity biomechanics during walking in boys who are overweight versus healthy weight. Pediat Phys Ther 2009;21:187–93. [14] Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic,kinetic, and electromyographic data in normal adult gait. J Orthop Res 1989;7:849–60. [15] Manal K, McClay I, Stanhope S, Richards J, Galinat B. Comparison of surface mounted markers and attachment methods in estimating tibial rotations during walking: an in vivo study. Gait Posture 2000;11:38–45.
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Sagittal and frontal plane joint mechanics throughout the stance phase of walking in adolescents who are obese A.G. McMillan a,*, A.M.E. Pulver a, D.N. Collier b, D.S.B. Williams a a b
Department of Physical Therapy, 2410E Health Sciences Building, East Carolina University, Greenville, NC 27834, United States Department of Pediatrics, Brody School of Medicine at East Carolina University, Greenville, NC, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 September 2009 Received in revised form 25 March 2010 Accepted 8 May 2010
The incidence of obesity has increased dramatically in children and adolescents, and with this comes health risks typically associated with adult obesity. Among those health consequences are musculoskeletal damage and pain. Previous studies have demonstrated inconsistent effects of increased body mass on movement patterns in adults and children who are obese. The purpose of this study was to investigate frontal and sagittal plane mechanics during walking in adolescents who were obese. Adolescents (12–17 years) who were obese were recruited from a weight management program, and healthy weight peers (matched for age, race and gender) were recruited from the community. Three-dimensional motion analysis of the lower extremities was performed during walking. Analysis of kinematic and kinetic data from 36 adolescents who were obese and healthy weight revealed significant differences in mechanics at all lower extremity joints in both sagittal and frontal planes. Subjects who were obese seemed to use movement strategies that minimized joint moments, especially at the hip and knee during walking. The lower extremity mechanics during walking in the subjects who were obese raise concerns about maintenance of structural integrity of the lower extremity joints over time, given the repeated high stresses across the joints even with walking. Neither the long term consequences of these atypical movement patterns, nor the ability to alter these patterns through therapeutic activities or weight loss has been investigated in adolescents who are obese. ß 2010 Elsevier B.V. All rights reserved.
Keywords: Obesity Adolescents Walking Biomechanics
1. Introduction Childhood obesity, defined by the Centers for Disease Control and Prevention as having a body mass index (BMI) greater than the 95th percentile for age and gender, has become an epidemic, with 17.4% of all adolescents (12–19 years) in the United States classified as obese [1]. Chronic conditions such as cardiovascular disease and diabetes, commonly seen in adults who are obese, are now seen in adolescents who are obese [2,3]. Musculoskeletal conditions, including osteoarthritis, low back pain, and soft tissue injury, are often associated with obesity [4,5]. Increased body mass, with increased forces across weight-bearing joints, has been causally implicated in many of these musculoskeletal conditions [5]. Forces on joint surfaces are increased during any weightbearing activity, including walking. Increased body mass may increase risk of damage and injury to joint surfaces and other musculoskeletal structures with repetitive loading during weightbearing activities. Many have hypothesized that movement patterns are significantly different in individuals who are obese.
* Corresponding author. Tel.: +1 252 744 6232; fax: +1 252 744 6240. E-mail address: [email protected] (A.G. McMillan). 0966-6362/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2010.05.008
While most studies report that children and adults who are obese walk slower, with a wider step width, shorter steps, and increased double support time/decreased single limb support time [6–9]. Nantel et al. [10] reported no difference in gait velocity, cadence, stride length or double limb support time in obese versus healthy weight (HW) children. Reports of kinematic and kinetic characteristics of walking in individuals who are obese are inconsistent. Peak knee flexion angles during stance phase of walking have been reported as being lower in children and adults who are obese [11,12] while others reported no difference in these angles [8,9]. Peak plantarflexion angles have been reported as lower [9] and higher [10] in obese versus HW subjects. The lower plantarflexion angles were noted in subjects with significantly slower gait velocity [9]. No differences in sagittal plane hip and knee moments have been found between obese and HW subjects during walking [8,11,12]. Compared to HW adults, sagittal plane ankle moments of obese adults have been reported as lower [8] and higher [12] during walking. Nantel et al. [10] reported the only difference in lower extremity sagittal plane kinetics during walking was an earlier shift from hip extension moment to hip flexion moment in obese versus HW children. Less data is available for frontal plane biomechanics during walking. Compared to HW individuals, peak hip frontal plane
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
264
angles during stance phase have been reported as significantly more abducted in adults who are obese [9], and as significantly more adducted in children who are obese [13]. The greater hip abduction angles were reported in subjects who walked significantly slower than their lean peers [9]. Greater knee valgus and rearfoot inversion during walking has been reported in children who are obese (compared to HW peers) [13], while studies have alternately reported higher knee adduction moments [13] and knee abduction moments [11] in obese versus HW children. In summary, a variety of investigations of the effects of obesity on characteristics of walking have yielded inconsistent results. Studies to date have included adults and children (8–13 years old). Frontal and sagittal plane biomechanics in hip, knee and ankle at specific points during the stance phase of walking in adolescents who are obese have not been reported. The purpose of this study was to examine the sagittal and frontal plane lower extremity biomechanics during walking in adolescents who were obese versus HW peers. 2. Methods 2.1. Subjects Both male and female adolescents who were obese and HW were recruited for this study. Inclusion criteria were: age (12–17 years), BMI (less than the 85th percentile for age and gender for HW group; and greater than 95th percentile for age and gender for obese group). Potential subjects were excluded if they had musculoskeletal, neuromuscular, and/or cardiopulmonary conditions other than obesity that would limit movement or compromise safety. Participants who were obese were recruited from a local healthy weight program. Healthy weight participants were recruited from the community and matched by age, gender, and race. Written informed consent by a parent/legal guardian and written assent by the adolescent were obtained for each participant prior to data collection. This study was approved by the University and Medical Center Institutional Review Board. 2.2. Equipment Kinematic data were collected using eight Hawk infrared digital cameras (Motion Analysis Corp., Santa Rosa, CA) with a sample rate of 120 Hz. Force data were collected using two synchronized force plates (AMTI, Watertown, MA) with a sampling rate of 960 Hz. Retro-reflective spherical markers were attached directly to the skin on each subject’s trunk and lower extremities at the sternum, T1, T10, L5/ S1, iliac crests, acromion processes, greater trochanters, medial and lateral femoral condyles, medial and lateral malleoli, posterior and lateral heels, and 1st and 5th metatarsal heads. Rigid arrays of markers were also placed on thighs and lower legs. Anatomical/joint markers (those defining joint centers and segment coordinate axes) were left on only during static calibration trial; 22 tracking markers remained on the subject’s calcaneus, shank, thigh, pelvis, and trunk during walking trials [13]. 2.3. Experimental procedures Participants were instructed to walk down a 15-foot walkway at a self-selected pace without looking down at the ground. In order to account for any differences in lower extremity mechanics that may exist due to differences in gait velocity, subjects were further matched on gait velocity. This resulted in 18 subjects in each group (Table 1). Each participant performed warm-up trials prior to data collection to establish a starting point so that each foot hit one of the force platforms during walking trials. Data collection continued until the subject completed at least 10 successful trials in which each foot contacted a force platform with no apparent change in gait pattern to purposefully hit the platforms.
Table 2 Sagittal plane kinematic and kinetic variables, with mean (SD) and p values. HW: healthy weight. Obese Sagittal plane kinematic variables (8) Ankle angle at initial 0.87 contact Ankle peak angle during 7.77 early stance Ankle peak angle during 7.19 late stance Ankle angle at toe off 9.99 Knee angle at initial contact 1.38 Knee peak angle during 11.26 early stance Knee peak angle during 4.17 midstance Knee angle at toe off 40.82 Hip angle at initial contact 18.01 Hip peak angle during 12.45 late stance Hip angle at toe off 7.76
HW
p-Value
(6.01)
0.76 (2.77)
0.15
(5.32)
4.69 (2.21)
0.02
(5.17)
7.18 (3.31)
0.50
(5.08) (7.35) (7.03)
6.82 (4.09) 7.10 (3.41) 16.09 (6.35)
(7.93)
7.22 (4.72)
0.02 0.003 0.02 0.09
(7.90) (10.50) (11.04)
37.28 (5.28) 30.47 (9.62) 4.69 (9.90)
(11.73)
1.00 (9.61)
0.01
0.11 (0.04)
0.44
Sagittal plane kinetic variables (Nm/kg m) Ankle peak moment during 0.11 (0.05) early stance Ankle peak moment during 0.67 (0.13) late stance Knee peak moment at initial 0.19 (0.06) contact Knee peak moment during 0.20 (0.14) early stance Knee peak moment during 0.10 (0.14) late stance Hip peak moment at 0.43 (0.12) initial contact Hip peak moment during 0.37 (0.16) late stance
0.06 0.000 0.02
0.88 (0.07)
0.000
0.28 (0.11)
0.001
0.12 (0.14)
0.05
0.31 (0.11)
0.000
0.72 (0.23)
0.000
0.24 (0.08)
0.002
gait (0% = initial contact, 100% = toe off). For subjects who were obese, hip joint centers were defined using a modified model to account for excessive subcutaneous tissue that would have significantly offset the location of the hip joint centers in these subjects. Specifically, the hip joint centers were calculated by first placing the anatomical markers on the skin over the greater trochanters. Anthropometric measures were then taken with calipers to approximate how much soft tissue was over the greater trochanters. This number was then used as an offset number and it was entered into the model file for Visual 3D. Virtual greater trochanter markers were then established and used to determine hip joint centers. Finally, hip joint centers were estimated by taking 25% of distance from each virtual greater trochanter marker. Moments were normalized to subject’s mass and height. Sagittal and frontal plane angles and moments at specific events (initial contact, toe off), and peak angles and moments during early stance (1st 30%), midstance (40–60%), and late stance (last 30%) were selected for statistical analysis. Data were averaged across all trials for each subject, and group averages were calculated and used for statistical analyses. Student’s t-tests were used to determine the differences between the obese and HW groups on the variables of interest. An a value of 0.05 was selected for statistical significance. Bonferroni corrections were performed due to multiple comparisons, resulting in corrected p value of 0.005 for kinematic data and 0.007 for kinetic data.
2.4. Data analysis
3. Results
Raw coordinate data were smoothed using a second order recursive Butterworth filter at 12 Hz for kinematics and 50 Hz for kinetics. EvaRT v5.0.4. Software (Motion Analysis Corp., Santa Rosa, CA) was used for data collection and initial processing of the raw data. Further processing was performed using Visual 3D software (C-Motion, Inc., Germantown, MD). Trials were normalized to the stance phase of
Sagittal plane. Subjects who were obese had a significantly lower plantarflexion moment during late stance compared to the HW group (Table 2 and Fig. 1D). The obese group had significantly less knee flexion at initial contact (Table 2 and Fig. 1B). In general
Table 1 Subjects characteristics.
Obese group (n = 18) HW group (n = 18) p-Value
Female/Male
African-American/ Caucasian
Age (years)
Height (m)
Weight (kg)
Actual BMI
BMI % (for age/gender)
BMI z-score
17/1 13/5
15/3 14/4
15.0 (1.5) 14.6 (1.8) 0.23
1.6 (0.1) 1.6 (0.1) 0.10
121.2 (30.8) 53.2 (7.0) 0.000
44.6 (10.2) 20.3 (2.0) 0.000
> 99 55.1 (22.8)
2.54 (0.34) 0.15 (0.65) 0.000
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
265
Fig. 1. Sagittal plane lower extremity kinematics (A–C) and kinetics (D–F) during stance phase of walking in adolescents who are obese (black) and healthy weight (HW, gray). Positive values are ankle dorsiflexion, knee extension, hip flexion. Solid lines = mean, dotted lines = 1SD.
the subjects who were obese maintained a relatively more extended knee throughout stance. Knee flexion moment was significantly lower in the obese group at initial contact and in late stance (Table 2 and Fig. 1E). At the hip, subjects who were obese remained less flexed/more extended than the HW group throughout stance, but this difference was only significant at initial contact (Table 2 and Fig. 1C). Subjects who were obese had significantly lower hip extension moments at initial contact, significantly higher hip flexion moments during late stance, and exhibited an earlier transition to hip flexion moment, compared to the HW group (Table 2 and Fig. 1F). Frontal plane. Subjects who were obese had a significantly lower rearfoot inversion moment during late stance, compared to their HW peers (Table 3 and Fig. 2D). At the knee, HW subjects had a neutral to adducted knee position, while subjects who were obese maintained an abducted (valgus) knee position throughout stance (Table 3 and Fig. 2B). Peak frontal plane knee angles were significantly different between groups at initial contact, during
early stance, and during late stance (Table 3 and Fig. 2B). Subjects who were obese had significantly greater frontal plane excursion between peak knee angle during early stance and peak angle during late stance compared to their HW peers (Table 3 and Fig. 2B). Frontal plane knee moments were also significantly different between groups, with subjects who were obese exhibiting lower knee abduction moments throughout much of stance (significant during all phases of stance; Table 3 and Fig. 2E). Hip angles were not significantly different between groups, with both groups exhibiting adduction then abduction during stance (Table 3 and Fig. 2C). The obese group did exhibit a significantly lower hip abduction moment during early stance (Table 3 and Fig. 2F). 4. Discussion The purpose of this study was to investigate the sagittal and frontal plane lower extremity biomechanics during walking in adolescents who were obese versus adolescents who were HW.
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
266
Table 3 Frontal plane kinematic and kinetic variables, with mean (SD) and p values. HW: healthy weight. Obese Sagittal plane kinematic variables (8) Rearfoot angle at 10.57 (3.92) initial contact Rearfoot angle peak 1.96 (3.47) midstance Rearfoot excursion 9.88 (2.69) midstance ! late stance Knee angle at 4.12 (4.37) initial contact Knee angle peak 0.76 (4.32) during early stance Knee excursion initial 3.36 (1.70) contact ! early stance Knee angle peak during 9.55 (7.62) late stance Knee excursion 8.79 (4.49) loading ! late stance Hip angle peak during 7.00 (3.47) early stance Hip angle excursion initial 6.12 (2.85) contact ! early stance Hip angle peak during 7.06 (3.80) late stance Frontal plane kinetic variables (Nm/kg m) Rearfoot moment amplitude 0.03 during early stance 0.07 Rearfoot moment peak amplitude during late stance Knee moment peak amplitude 0.03 during early stance Knee moment peak amplitude 0.16 during midstance Knee moment peak amplitude 0.14 during late stance Hip moment peak amplitude 0.42 during early stance Hip moment peak amplitude 0.47 during late stance
HW
p-Value
8.26 (4.88)
0.13
0.27 (5.03)
0.25
10.73 (3.34)
0.49
0.35 (2.46)
0.003
2.99 (2.73)
0.004
3.35 (1.35)
0.98
0.80 (3.94)
0.000
3.79 (2.90)
0.000
6.72 (4.20)
0.89
6.57 (2.95)
0.65
5.44 (3.18)
0.18
(0.01)
0.02 (0.02)
0.77
(0.03)
0.11 (0.02)
0.000
(0.03)
0.07(0.06)
0.003
(0.06)
0.30 (0.09)
0.000
(0.06)
0.27 (0.09)
0.000
(0.12)
0.55 (0.13)
0.003
(0.19)
0.54 (0.09)
0.18
Significant differences between groups were found in sagittal and frontal plane kinematics and kinetics at all three lower extremity joints during walking. Sagittal plane. The only significant difference between groups at the ankle in the sagittal plane was the peak plantarflexion moment in late stance, which was lower in the obese group. Other reports of sagittal plane ankle kinematics and kinetics in obese (versus HW) subjects have been inconsistent: no differences in motion, but lower muscle moments at the ankle [8]; lower plantarflexion angles and greater dorsiflexion angles [9]; greater plantarflexion motion and higher ankle moments [12]; lower plantarflexion moments [11]. Lower plantarflexor moments seen in this and other studies might be due to relative weakness, as the plantarflexors are partial contributors to ankle joint moments in the sagittal plane. Lower plantarflexion moment in late stance results in decreased push-off which will likely require increased contributions at other joints. The increased hip flexion moment seen in late stance in the obese subjects in this study is likely a compensation to pull the limb into swing rather than push it through with the plantarflexors. Several significant findings were noted in the sagittal plane at the knee joint. In general, subjects who were obese had less flexion during early stance and lower flexion moments at the knee at initial contact and during late stance. The knee angle findings are similar to those previously reported [11,12], though others reported no significant differences in knee sagittal plane angles [8,9]. Subjects who are obese may keep the knee less flexed to
compensate for potential knee extensor weakness, or less effective knee extensor pull given their more abducted knee alignment in stance. A relatively extended knee during stance may also be a compensation for instability within the knee joint structure. Neither knee extensor strength nor knee stability were directly measured in this study. Subjects who were obese had significantly less hip flexion and lower hip extension moments at initial contact, and higher hip flexion moment during late stance, compared to the HW group. Others have reported no difference in sagittal plane kinematics and kinetics at the hip [8,9,12]. Obese subjects in the current study also moved much earlier (during early stance) into hip extension with a concurrent hip flexion moment. This could be a compensation for relatively weak hip extensors: extension at the hip during stance brings the line of force closer to or behind the hip joint, creating an external extensor moment at the hip and thus requiring subjects to use less hip extensor strength during stance. Hip extensor strength was not reported in this study. Frontal plane. Unlike previous findings [13], rearfoot kinematics in the frontal plane were not different between groups, perhaps due to inclusion of females in the current study. Further investigation of differences in mechanics between males and females who are obese is warranted. Subjects who were obese did exhibit significantly lower rearfoot inversion moment in late stance, which may be due to relative weakness in the muscles controlling rearfoot eversion. While rearfoot motion was not different between groups, even a slight change in the lateral position of the ground reaction force relative to the joint center would likely explain changes in joint moments as reported here. As reported previously [13], subjects who were obese maintained an abducted (valgus) position at the knee throughout stance, while HW subjects remained neutral/adducted at the knee. Subjects who were obese also had significantly greater frontal plane knee excursion during stance. While in a previous study boys who were obese exhibited knee adduction moments during early and late stance and knee abduction moments through midstance, in the current study obese subjects’ knee moments remained in abduction throughout stance, and were significantly lower than those of HW subjects. Subjects in Gushue et al. [11] also exhibited knee abduction moments, though these moments were greater in subjects who were obese (versus HW). Differences in the current results may be partially due to the model used to calculate hip joint centers being based on actual anthropometric measurements. The repetitive stresses on the knee joint structures related to the greater frontal plane excursion and moments during stance is concerning due to the potential for damage to knee joint structures, pain, limited motion, and resultant disability. Frontal plane hip angles were not different between groups, and only the hip abduction moment during early stance was significantly different between groups. Differences from previous data [13] may be partially due to calculation of hip joint centers. Spyropoulos et al. [9] reported that adults who were obese maintained hip abduction during stance, but those data were based on two-dimensional skin markers (thus not taking into account the amount of subcutaneous tissue present in obese subjects). The lower hip abduction moment during early stance suggests relative hip abductor weakness, though data is not available at this time to confirm this suggestion. 5. Limitations Both males and females were recruited for this study in order to improve the generalizability of the results. While combining male and female subjects does introduce a source of potential variability, there were too few males actually enrolled in the
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
267
Fig. 2. Frontal plane lower extremity kinematics (A–C) and kinetics (D–F) during stance phase of walking in adolescents who are obese (black lines) and healthy weight (HW, gray lines). Positive values are rearfoot eversion, knee adduction, hip adduction. Solid lines = mean, dotted lines = 1SD.
current study to compare them separately. Potential differences in movement characteristics between males and females who are obese will need further study. Skin movement and marker placement errors are potential confounders of movement data, especially in subjects who are obese. Several methods were employed in the current study in an attempt to minimize these errors. The same investigator placed all markers on all participants to account for differences in marker placement due to investigators or days [14]. To limit skin and tissue motion, markers were attached to rigid shells, which were in turn attached to tight circumferential wraps on the distal portion of the tibia and femur. This method and location of attachment has been shown to minimize soft tissue motion relative to bone [15]. The choice was made not to use tight fitting shorts or other material around the hip to try to decrease this marker movement because the binding or stabilizing effect of the material itself could lead to alterations in gait characteristics. We are confident that the differences observed between groups were minimally due to skin movement or to marker placement.
6. Summary Subjects who were obese exhibited significant differences in both sagittal and frontal kinematics and kinetics of the lower extremities during the stance phase of walking. This was especially true at the knee, a joint which is prone to damage and injury. We hypothesize that muscle weakness is one potential cause of these movement differences, and are currently investigating the correlations between muscle strength at the hip and knee and the observed gait deviations in this group of subjects. An additional investigation is focusing on resistive strength training of lower extremity muscles implicated in the observed gait pattern. Because these movement differences are noted even in a relatively low stress movement such as walking, and because adolescents who are obese must walk on a daily basis (and in fact are often encouraged to walk to increase their physical activity), it is important to determine the true mechanisms behind this atypical gait pattern, and how/if these mechanisms can be addressed clinically. The long term consequences of atypical gait patterns in obese adolescents have yet to be identified.
268
A.G. McMillan et al. / Gait & Posture 32 (2010) 263–268
Acknowledgements This study was funded in part by a Research Development Award to Dr. Gross McMillan from East Carolina University. These funds were used for supplies, subject reimbursement, and research assistant salary. Conflict of interest statement None of the authors on this manuscript had or have any financial and personal relationships with other people or organizations that could inappropriately influence (bias) this work. References [1] Centers for Disease Control and Prevention. Overweight and obesity: childhood overweight. http://www.cdc.gov/nccdphp/dnpa/obesity/childhood/ index.htm. Accessed August 24, 2008. [2] Patterson RE, Frank LL, Kristal AR, White E. A comprehensive examination of health conditions associated with obesity in older adults. Am J Prev Med 2004;27:385–90. [3] Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord 1993;23:S2–11.
[4] Kortt M, Baldry J. The association between musculoskeletal disorders and obesity. Aust Health Rev 2002;25:207–14. [5] Anandacoomarasamy A, Caterson I, Sambrook P, Fransen M, March L. The impact of obesity on the musculoskeletal system. Int J Obes 2008;32:211–22. [6] Hills AP, Parker AW. Gait characteristics of obese children. Arch Phys Med Rehabil 1991;72:403–7. [7] Hills AP, Parker AW. Locomotor characteristics of obese children. Child Care Health Dev 1992;18:9–34. [8] Browning RC, Kram R. Effects of obesity on the biomechanics of walking at different speeds. Med Sci Sports Exerc 2007;39:1632–41. [9] Spyropoulos P, Pisciotta JC, Pavlou KN, Cairns MA, Simon SR. Biomechanical gait analysis in obese men. Arch Phys Med Rehabil 1991;72:1065–70. [10] Nantel J, Brochu M, Prince F. Locomotor strategies in obese and non-obese children. Obesity 2006;14:1789–94. [11] Gushue DL, Houck J, Lerner AL. Effects of childhood obesity on three-dimensional knee joint biomechanics during walking. J Pediatr Orthop 2005;25: 763–8. [12] DeVita P, Hortobagyi T. Obesity is not associated with increased knee joint torque and power during level walking. J Biomech 2003;36:1355–62. [13] McMillan AG, Auman NL, Collier DN, Williams DSB. Frontal plane lower extremity biomechanics during walking in boys who are overweight versus healthy weight. Pediat Phys Ther 2009;21:187–93. [14] Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic,kinetic, and electromyographic data in normal adult gait. J Orthop Res 1989;7:849–60. [15] Manal K, McClay I, Stanhope S, Richards J, Galinat B. Comparison of surface mounted markers and attachment methods in estimating tibial rotations during walking: an in vivo study. Gait Posture 2000;11:38–45.