coordinationpatternofthethigh,pelvic,andlumbar

Research ArticleCoordination Pattern of the Thigh Pelvic and LumbarMovements during the Gait of Patients with Hip Osteoarthritis

Takuya Ibara12MasayaAnan3 RyosukeKarashima2 KiyotakaHada2 Koichi Shinkoda45

Mahito Kawashima6 and Makoto Takahashi 45

1Graduate School of Biomedical and Health Sciences Hiroshima University 2-3 Kasumi 1-Chome Minami-kuHiroshima 734-8553 Japan2Department of Rehabilitation Kawashima Clinic 11-1 Miyabu Nakatsu Oita 871-0012 Japan3Physical erapy Course Faculty of Welfare and Health Science Oita University 700 Dannoharu Oita 870-1192 Japan4Department of Biomechanics Graduate School of Biomedical and Health Sciences Hiroshima University 2-3 Kasumi 1-ChomeMinami-ku Hiroshima 734-8553 Japan5Center for Advanced Practice and Research of Rehabilitation Graduate School of Biomedical and Health SciencesHiroshima University 2-3 Kasumi 1-Chome Minami-ku Hiroshima 734-8553 Japan6Department of Orthopaedic Kawashima Orthopaedic Hospital 17 Miyabu Nakatsu Oita 871-0012 Japan

Correspondence should be addressed to Makoto Takahashi biomechiroshima-uacjp

Received 27 February 2020 Revised 4 July 2020 Accepted 6 July 2020 Published 24 July 2020

Academic Editor Hongbo Zhang

Copyright copy 2020 Takuya Ibara et al is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

ere are limited reports on segment movement and their coordination pattern during gait in patients with hip osteoarthritis Toavoid the excessive stress toward the hip and relevant joints it is important to investigate the coordination pattern between thesesegment movements focusing on the time series datais study aimed to quantify the coordination pattern of lumbar pelvic andthigh movements during gait in patients with hip osteoarthritis and in a control group An inertial measurement unit was used tomeasure the lumbar pelvic and thigh angular velocities during gait of 11 patients with hip osteoarthritis and 11 controls evector coding technique was applied and the coupling angle and the appearance rate of coordination pattern in each directionwere calculated and compared with the control group Compared with the control group with respect to the lumbarpelvicsegment movements the patients with hip osteoarthritis spent more rates in anti-phase and lower rates in in-phase lateral tiltmovement With respect to the pelvicthigh segment movements the patients with hip osteoarthritis spent more rates within theproximal- and in-phases for lateral tilt movement Furthermore patients with osteoarthritis spent lower rates in the distal-phasefor anteriorposterior tilt and rotational movement Patients with hip osteoarthritis could not move their pelvic and thighsegments separately which indicates the stiffness of the hip joint e rotational movement and lateral tilt movements especiallywere limited which is known as Duchenne limp To maintain the gait ability it seems important to pay attention to thesedirectional movements

1 Introduction

Patients with hip osteoarthritis (OA) often have altered gaitpatterns which reportedly may accelerate disease progres-sion and severity or worsen symptoms [1ndash3] Altered gaitpatterns involving unusual movements in the hip joint causeother joints to move excessively to compensate for the hipjoint ese compensational movements may put additionalstress on these joints leading to dysfunction [4 5]

represented as coxitis knee [6] or hip-spine syndrome [5 7]It can be assumed that the cause of these compensationalmovements is the hip stiffness Steinhilber et al [8] showedthat patients with hip OA are likely to report the complaintof hip stiffness Tateuchi et al [9] reported that patients withhip OA who had undergone total hip replacement (THR)were likely to experience dynamic hip joint stiffness whereasFoucher et al [10] reported that patients were likely to retaintheir preoperative gait pattern after THR Although the

HindawiJournal of Healthcare EngineeringVolume 2020 Article ID 9545825 9 pageshttpsdoiorg10115520209545825

dynamic range of motion of the hip during gait had re-portedly improved among post-THR patients it was notcomparable with that in normal individuals [10] Otherstudies also examined the gait of patients with hip OA whohad not undergone THR by studying joint kinematics ki-netics and their correlations [4 11ndash15] But it remainsunclear whether the hip joint stiffness is present in the gait ofpatients with hip OA preoperatively Traditional measure-ments such as examining the range of motion during gaitare not enough for understanding control mechanisms It isimportant to examine the gait and movement patterns ofpatients with hip OA by focusing on time interval such asjoint coordination [16] e vector coding technique (VCT)allows researchers to examine the coordination pattern ofthe movement over time [17 18] Several studies found thatthe people with pain or musculoskeletal disorders such aspatellofemoral pain low back pain or idiopathic scoliosisshowed less movement between two relevant segments orjoints using the VCT [18ndash20] Smith et al [21] investigatedthe coordination pattern between pelvis and thigh with weakhip muscle in healthy subjects and Samaan [22] studied thecoordination pattern between hip and knee movement inhip OA patients with VCTtechnique However there was nostudy that investigated the thighndashpelvisndashlumbar coordina-tion pattern of patients with hip OA preoperatively eVCT characterizes the movement coordination pattern oftwo segments or joints as the predominance of movementspeed using the following four phases [16 23] in-phase (thesame directional movement with similar movement speed)anti-phase (opposite directional movement with similarmovement speed) proximal-phase (relatively large proximaljointsegment movement speed) and distal-phase (relativelylarge distal jointsegment movement speed) e ldquoin-phaserdquomovement especially indicates a particular movementpattern with similar movement speed between two segmentsor joint [16 23] e ldquoin-phaserdquo and ldquostiffrdquo are not actuallyequal in point of considering the force applied former re-ferring to the observed movement only and latter referringto the movement as the response to the force appliedHowever it has been conceivable that the in-phase coor-dination pattern relates to the stiffened condition [20] Webelieve that using the VCT to directly assess the patterns ofmovement for patients with hip OA will give us theknowledge regarding the stiff hip movement during thepelvicthigh ldquoin-phaserdquo coordination pattern as well as therelationship between the hip and relevant region movementsover time in patients with hip OA

We measured changes in segmental movement using aninertial measurement unit (IMU) e IMU is inexpensiveand easy to use in a clinical setting because the IMU does notneed the laboratory setting which is needed for using motioncapture system Other groups have documented the reli-ability and validity of using an IMU for measuring gaitkinematics [24ndash26] and assessing the symptoms of mus-culoskeletal disorders [15 25 27 28] In the first step ofVCT the change in the angles of two segmentaljointmovements can be calculated e procedure thus enablesthe calculation of the angular velocity which is the differencebetween two successive angles us we decided to use this

procedure in addition to collecting angular velocity datausing IMU

e aim of this study was to investigate the coordinationpattern of segmental movements in the gait of patients withhip OA using VCTand IMU which allows us to suggest thepresence of stiffness and the relationship of each segmentalmovement We hypothesized that patients with hip OA willreveal increased rate of ldquoin-phaserdquo movement pattern due tostiffness in the hip joint We also hypothesized that the in-phase coordination pattern of pelvicthigh movementswould have an anteriorposterior tilt due to limited hipextension movements among patients with hip OA

2 Materials and Methods

21 Participants e present study included 11 patientswith hip OA from an outpatient clinic and 11 age- andgender-matched community-dwelling control volunteerse inclusion criteria of the hip OA group were as followsbeing female ability to walk without the aid of medicalequipment age 40ndash70 years and having a grade of 3 or 4 onthe Kellgren-Lawrence grading system ere were very fewmale patients and the presence of differences with respect topatient sex was anticipated therefore we did not includemale patients in the study Patients with neurologicalconditions and other lower extremity joint disorders whichaffected the required tasks were excluded from the studyAmong patients with bilateral hip OA the more affected sidewas chosen as the ldquotargetedrdquo side e control group in-cluded volunteers who did not have any complaint abouttheir hips e other criteria were the same as those for hipOA group Data on study participants are found in Table 1

All procedures were approved by the institutional ethicscommittee and all participants provided written informedconsent prior to participating in this studye experimentalprocedures of the study were conducted according to theDeclaration of Helsinki

22 Data Collection ree IMUs (TSND151 ATR-pro-motions Soraku Japan) each equipped with a triaxial ac-celerometer gyroscope andmagnetometer were attached tothe following areas on the lateral side of the thigh betweenthe femoral condyle and greater trochanter (thigh) on thedorsal side between both posterior superior iliac spines(pelvis) and on the dorsal side at the first lumbar spinousprocess (lumbar) Sensor axes were aligned to carefullydetect the anterior-posterior medial-lateral and superior-inferior directions of movements e sampling rate was setat 100Hz

e participants were required to perform a uniform setof tasks First the participants were asked to stand quietlyAfter standing quietly they were directed to begin walkingover a distance of 30m following an oral cue by the ex-aminer Participants were asked to walk the path until theexaminer gave them an oral cue to stop Once directed tostop walking participants were required to maintain a quietstanding posture until given an oral cue by the examinerParticipants did not wear shoes during these tasks and were

2 Journal of Healthcare Engineering

asked to maintain a comfortable gait speed while performingthe tasks

23 Data Analysis e angular velocity and accelerationrates of each participantrsquos gait were recorded as theycompleted the task e initial and final moments of contactfor the target limb were identified using the recorded pelvisvertical acceleration data according to a method found in aprevious report [29] e thigh anteriorposterior tilt an-gular velocity data was used to distinguish between the leftand right initial contact Except for the first two gait cyclesfifteen successive stance phases of gait cycles were furtheranalyzed for this study

e time of each stance phase was normalized to a 100-point data scale as reported previously [17] To distinguishbetween the lumbarpelvic and pelvicthigh coordinationpatterns the coupling angle (CA) was calculated using thedirectional angular velocities of each segment In this studythe thigh external rotation posterior tilt (flexion) and lateraltilt (adduction) were considered positive values Regardingpelvic and lumbar movements rotation to ipsilateral side oftarget limb posterior tilt and lateral obliquity toward theipsilateral side of the target limb were considered positivevalues

In the original report [17] the CA was calculated usingthe arctangents of the differences in angle data duringsuccessive two instants of two segments Because the dif-ference in angle data at successive two instants is the angularvelocity vector we used the angular velocity data recordedfrom IMU to calculate CA as follows

ci tanminus1 αi

βi

1113888 1113889 (1)

where c was the CA calculated from the distal segmentangular velocity (α) and the proximal segment angularvelocity (β) at time i

e calculated CA at each time point was then averagedover the 15 stance phases For calculating average CA theCA vectors were divided into horizontal (x) and verticalcomponents (y) and then the averaged CA (c) was cal-culated is procedure was adjusted within subject asfollows

xi 1n

1113944

n

i1cos ci( 1113857 (2)

yi 1n

1113944

n

i1sin ci( 1113857 (3)

ci

arctanyi

xi

1113888 1113889 xi gt 0

180 + arctanyi

xi

1113888 1113889 xi gt 0

⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(4)

e averaged CA within subjects was classified at eachinstant according to the following four conditions proxi-mal-phase in-phase distal-phase and anti-phase [16 23](Figure 1) It was expected that the ldquoin-phaserdquo pattern wasassociated with the presence of stiffness because theproximal and distal segments tend to move together eappearance rates that is the occupancy rate of the targetphase during stance phase of classified coordination pat-terns during the averaged CA were used for statisticalanalysis To determine the times series data for the typicalCA within each group we also calculated the averaged CAwithin groups using the same procedure described aboveFurthermore we calculated the coupling angle variability(CAV) as the index of variability [16] which was calculatedby the length of averaged coupling angle (δi) as follows

δi

xi2

+ yi2

1113969

(5)

CAVi

2 middot 1 minus δi1113872 1113873

1113969

middot180π

(6)

Additionally we calculated the maximum value ofsegment angular velocity in each direction of each segmentduring stance phase by calculating the average of themaximum values in each stance phase

Data Analysis was performed with using MATLAB2017a (MathWorks Natick MA USA)

24 Statistical Analysis e anthropometric data wascompared with independent-sample t-test Welchrsquos t-test orMannndashWhitney U test in accordance with the normality(ShapirondashWilk test) and homoscedasticity of the data eappearance rates for coordination patterns and the angularvelocity data between the control and hip OA groups werecompared with using the same process as the anthropo-metric data except for the use of the Bonferroni correctionFor the anthropometric data and the angular velocity datathe effect size was calculated using Cohenrsquos d Because thebody weight and the body mass index were significantlydifferent between the control group and the hip OA groupwe also performed the analysis of covariance (ANCOVA)with body mass index as covariate and calculated η2 as the

Table 1 Anthropometric data of control and hip OA groups

Parameter Control group Hip OA group p Effect sizeAge (year) 538 (69) 581 (74) 0178 0600Body mass (kg) 493 (51) 595 (83) 0002 1490Body height (m) 157 (005) 158 (006) 0583 0240Body mass index (kgm2) 201 (26) 238 (27) 0004 1380OA grade (IIIIV) mdash 56 mdash mdashData are represented as means (SD)

Journal of Healthcare Engineering 3

effect size We checked for group by body mass index in-teraction for each parameter since adjustments weremeaningful only if regression slopes were homogeneous Asa result we performed ANCOVA only for the parameters forwhich we found no evidence of interaction angular velocityof thigh lateral tilt and lumbar contralateral rotation esignificance level was set at 5 Statistical analyses wereperformed using SPSS 170 J for Windows (SPSS JapanTokyo Japan)

3 Results

31 Coordination Patterns Figure 2 shows the appearancerates of classified coordination pattern of lumbarpelvic andpelvicthigh of both groups Regarding the lateral tiltmovement of the lumbarpelvic region (Figure 2(a)) therewere significant differences between the in-phase and anti-phase appearance rates Patients in the hip OA group spentmore rates in anti-phase and lower rates in in-phase com-pared with control group ere were no significant dif-ferences in anteriorposterior tilt and rotational movementat lumbarpelvic region (Figures 2(b) and 2(c))

With regard to the pelvicthigh there were significantdifferences between the OA and control groups regardingthe proximal-phase and in-phase appearance rates for lateraltilt movement (Figure 2(d)) distal-phase rate for anteriorposterior tilt movement (Figure 2(e)) and distal-phase ratefor rotational movement (Figure 2(f)) Participants in thehip OA group spent more rates in the proximal- andin-phases for lateral tilt movement and lower rates in thedistal-phase rate for anteriorposterior tilt and rotationalmovement compared with participants in the control group

32 Time Series Data of CA and CAV e CA and CAVmeasurements for the control and hip OA groups are foundin Figure 3 In regard to the lateral tilt movements of thelumbarpelvic segments among patients with OA(Figure 3(a)) there was considerable increase in anti-phaseand significant decrease in the time of in-phase movementsafter the initial contact of stance phase (0ndash20) e CAabout the anteriorposterior tilt (Figure 3(b)) of the OAgroup did not decrease during the middle third of the stancephase (40ndash60) Furthermore the CA about rotationmovement remained constant during the first half of thestance phase (20ndash40) (Figure 3(c))

Among patients with OA (Figure 3(d)) there was aconsiderable increase during time spent in in-phase andproximal-phase pelvicthigh movements about the lateral tiltduring middle one-third of the stance phase (30ndash60)ere was a considerable decrease in the distal-phasemovement about the anteriorposterior tilt (Figure 3(e))during 70ndash80 of the stance phase and the decrease in thedistal-phase movement about rotation (Figure 3(f)) wasshown at 40ndash80 of stance phase e CAV was higher inthe hip OA group compared with control group except forthe lumbarpelvic and pelvicthigh rotational movements(Figures 3(c)ndash3(f))

33 Maximal Angular Velocity ere were significant dif-ferences between the control and OA groups regarding thighangular velocities about the medial tilt anterior tilt andposterior tilt (Table 2) e hip OA group showed lowerangular velocity in these movements compared with thecontrol group

4 Discussion

In this study the coordination patterns of the segmentalgroups were assessed using the CA which was calculatedusing the angular velocity data recorded using the IMU fromtwo segments during gait

We hypothesized that the hip OA patients would spend alot of rate in in-phase movement pattern particularly inanteriorposterior tilt movement at pelvicthigh segmentsHowever there was no significant difference in appearancerate of in-phase between both groups although it tended tobe high in hip OA group compared with control groupFurthermore the distal-phase appearance rate was signifi-cantly lower in the OA group compared with control groupese findings may imply that the patients with hip OArelied on pelvic anterior tilt rather than hip extension whenproducing the stride length as described in a previous study[4] and this may be caused by the hip joint stiffness aspreviously reported [9 13] Moreover this may be the causeof degenerative spondylolisthesis [30 31] Actually three outof 11 patients of the present hip OA group were diagnosedwith degenerative spondylolisthesisis findingmay also beconfirmed by the relationship between the smaller maximalangular velocity value of the thigh and insignificant dif-ference about maximal angular velocity of the pelvis ante-riorposterior tilt among patients in the hip OA group

Distal-phase

In-phase

In-phase

Coordination patternProximal-phaseln-phaseDistal-phaseAnti-phase

Coupling angle (γ) range (deg)0 le γ lt 225 1575 le γ lt 2025 3375 le γ le 360

225 le γ lt 675 2025 le γ lt 2475675 le γ lt 1125 2475 le γ lt 2925

1125 le γ lt 1575 2925 le γ lt 3375

Anti-phase

Anti-phase

Proximal segment angular velocity

(degs)

Distal segment angular velocity (degs)

Proximal-phaseCoupling

angle

Distal-phase

Proximal-phase

225

6751125

1575

2025

2475 2925

3375

Figure 1 e coupling angle between the resultant vector ofproximal and distal segments angular velocity data and the righthorizontal line e coupling angle was divided into four coor-dination patterns


Regarding lateral tilt movement the OA group spent ahigh rate in the in-phase and proximal-phase for pelvicthigh segment movements By contrast they spent a low ratein-phase and high rate in the anti-phase for lumbarpelvic

movements ese coordination patterns were observedwithin the first half of the stance phase with difference inlumbarpelvic coordination preceding the difference in thepelvicthigh coordination pattern (Figures 3(a) and 3(d)) It

80

70

60

50

40

30

20

10

0

Rate

()

lowast

lowast

Proximal Distal In AntiCoordination pattern

ControlHip OA

(a)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(b)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(c)

70

60

50

40

30

20

10

0

Rate

() lowast

lowast


ControlHip OA

(d)

100

80

60

40

20

0

Rate

()

lowast


ControlHip OA

(e)

70

60

50

40

30

20

10

0

Rate

()

lowast


ControlHip OA

(f )

Figure 2 e appearance rate of each phase at the lumbarpelvic and pelvicthigh segments about each directional movement e rate ofappearance of lumbarpelvic regions about lateral tilt (a) anteriorposterior tilt (b) and rotation movement (c) and that of pelvicthighregions about lateral tilt (d) anteriorposterior tilt (e) and rotation movement (f ) lowast plt 005


3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()

Hip OA coupling angleControl coupling angle

Hip OA coupling angle variability

Control coupling anglevariability

(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )

Figure 3 e time series data showing the mean coupling angle and coupling angle variability at lumbarpelvic and pelvicthigh segmentsabout each directional movement e mean coupling angle and coupling angle variability at lumbarpelvic regions about lateral tilt (a)anteriorposterior tilt (b) and rotation movement (c) and that at the pelvicthigh regions about lateral tilt (d) anteriorposterior tilt (e) androtation movement (f ) e gray circle indicates the mean coupling angle of the control group e white triangle indicates the meancoupling angle of the hip OA group e gray line indicates the coupling angle variability within the hip OA group e black line indicatesthe coupling angle variability within the control group


was previously thought that these patterns were a symptomof the Duchenne limp which was frequently seen in patientswith hip OA [15 32] e Duchenne limp is a movementperformed in order to compensate for weakened hip ab-ductor muscles [15 32] In this study it was demonstratedthat this movement pattern preceded the body weight-bearing increment (single support phase) Furthermore thisfinding means that the patients in the hip OA group mayadjust their movement pattern when anticipating the suc-cessive load increment to the hip joint or abductor musclesis finding could not recognize the maximal angular ve-locity because the data did not include time interval

In rotational movement a significant difference inmovement was only observed in the distal-phase appearancerate of the pelvicthigh region (Figure 2(f )) and the CA ofthe hip OA group was low during the last half of the stancephase (Figure 3(f )) In this movement the in-phase ap-pearance rate was not significantly different but the dif-ference was large in magnitude ese data may imply thatindividuals in the hip OA group were unable to move theirpelvic and thigh segments independently from each otherOne possible reason behind this stems from the instabilitycaused from a decrease in femoral head coverage andor thedemand to make the stride length not only the extensionmovement but also the rotation e result of a previousstudy [33] showed that the anterior femoral head coveragedecreased during late stance Most previous reports [13 14]showed that the range in hip extension movement amongpatients with hip OA decreased during the late stancePatients compensated for the decrease in coverage of theanterior femoral head via the anterior pelvic tilt [34] whichled to a decrease in the hip extension range e hip externalrotation (femoral external rotation relative to the pelvis)movement may also decrease the anterior coverage of thefemoral head during late stance us the decrease of theappearance rate of separated rotational movement at the

pelvicthigh segments indicates that the patients with hipOA fix their acetabular roof toward the femoral head bymaking their hip joint stiff

roughout all the directional movements in this studythe patients with hip OA seemed to move in ldquoin-phaserdquomovement pattern within their hips preoperatively bymaking their hip stiff by their own or by the contractureese findings can be obtained by examining the timecourse of joint coordination pattern during gait using VCTinstead of the conventional measurements such as therange of motion or angular velocity of each joint Withrespect to lumbarpelvic coordination the patients with hipOA moved their pelvis compensatory especially in lateraltilt movement e movements in anteriorposterior tiltand rotation may compensate within other segmentsjointssuch as thorax

is study had several limitations First the sample sizewas relatively small compared with previous studies Pre-vious reports suggested that not all patients of hip OA hadthe Duchenne limp and there were variations within thepatient pool [15] Secondly the symptoms of OA wererelatively severe among the patients who participated in thisstudy Because the movement coordination patterns may beinfluenced by the range of motion in the hip a larger samplesize and further subdivisions of groups may be necessaryMoreover thoracic movement was not measured in thisstudy e lumbar segment movement implies thoracicmovement especially in rotation and lateral tilt because oftheir anatomical structures But it remains unclear Finallythe angular velocity data collected in this study was notmeasured using a global coordinate system only a localcoordinate systemerefore it is important to take this intoaccount when comparing the results of this study with re-sults collected using a motion capture system Althoughthere were some limitations in this study the informationcollected regarding the coordination pattern of segment

Table 2 e maximal angular velocity data of each segment of control and hip OA groups

Segments Direction Control Hip OA p Effect size

Lumbar

Ipsilateral tilt (degs) 178 (140ndash294) 154 (140ndash176) 0200 0364Contralateral tilt (degs) 181 (59) 149 (33) 0129 0249Anterior tilt (degs) 216 (76) 207 (73) 0781 0063Posterior tilt (degs) 304 (107) 262 (92) 0341 0272

Ipsilateral rotation (degs) 411 (110) 409 (94) 0955 0016Contralateral rotation (degs) 375 (100) 296 (122) 0841 0002

Pelvis

Ipsilateral tilt (degs) 351 (85) 267 (117) 0069 0545Contralateral tilt (degs) 339 (317ndash359) 247 (208ndash345) 0071 0527Anterior tilt (degs) 222 (76) 227 (86) 0885 0035Posterior tilt (degs) 235 (70) 235 (99) 0996 0001


igh

Lateral tilt (degs) 537 (173) 411 (182) 0952 0001Medial tilt (degs) 905 (250) 498 (214) 0001 1944Anterior tilt (degs) 1089 (254) 824 (225) 0017 1252Posterior tilt (degs) 1602 (289) 1101 (369) 0002 1990

External rotation (degs) 1141 (247) 1044 (204) 0327 0470Internal rotation (degs) 688 (164) 769 (383) 0531 0343

Data are represented as means (SD) or medians (interquartile range) Bold texts indicate significant p values


movement will be helpful for improving physical therapyapproaches for patients with hip OA

5 Conclusion

is study investigated the coordination pattern of lumbarpelvis and thigh movements among patients with hip OAese patients showed less separated pelvicthigh move-ments compared with control participants suggesting stiff-ness of the hip joint However patients with OA also movedtheir lumbarpelvic regions in opposite directions in lateraltilt and in the same directions in rotational movementesemovements may decrease stress on the abductor musclesand provide stability to the hip joint

Data Availability

e data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

e authors declare that they have no conflicts of interest

References

[1] K C Foucher B R Schlink N Shakoor and M A WimmerldquoSagittal plane hip motion reversals during walking are as-sociated with disease severity and poorer function in subjectswith hip osteoarthritisrdquo Journal of Biomechanics vol 45no 8 pp 1360ndash1365 2012

[2] H Tateuchi Y Koyama and R Tsukagoshi ldquoAssociations ofradiographic degeneration and pain with daily cumulative hiploading in patients with secondary hip osteoarthritisrdquo Journalof Orthopaedic Research vol 34 no 11 pp 1977ndash1983 2016

[3] H Tateuchi Y Koyama H Akiyama et al ldquoDaily cumulativehip moment is associated with radiographic progression ofsecondary hip osteoarthritisrdquo Osteoarthritis and Cartilagevol 25 no 8 pp 1291ndash1298 2017

[4] E Watelain F Dujardin F Babier D Dubois and P AllardldquoPelvic and lower limb compensatory actions of subjects in anearly stage of hip osteoarthritisrdquo Archives of Physical Medicineand Rehabilitation vol 82 no 12 pp 1705ndash1711 2001

[5] D S Weinberg J J Gebhart and R W Liu ldquoHip-spinesyndrome a cadaveric analysis between osteoarthritis of thelumbar spine and hip jointsrdquo Orthopaedics amp TraumatologySurgery amp Research vol 103 no 5 pp 651ndash656 2017

[6] M Tsubosaka T Matsumoto K Takayama N Nakano andR Kuroda ldquoTwo cases of late medial instability of the kneedue to hip disease after total knee arthroplastyrdquo InternationalJournal of Surgery Case Reports vol 37 pp 200ndash204 2017

[7] P Ben-Galim T Ben-Galim N Rand et al ldquoHip-spinesyndromerdquo Spine vol 32 no 19 pp 2099ndash2102 2007

[8] B Steinhilber G Haupt R Miller S Grau P Janssen andI Krauss ldquoStiffness pain and hip muscle strength are factorsassociated with self-reported physical disability in hip oste-oarthritisrdquo Journal of Geriatric Physicalerapy vol 37 no 3pp 99ndash105 2014

[9] H Tateuchi R Tsukagoshi Y Fukumoto S Oda andN Ichihashi ldquoDynamic hip joint stiffness in individuals withtotal hip arthroplasty relationships between hip impairmentsand dynamics of the other jointsrdquo Clinical Biomechanicsvol 26 no 6 pp 598ndash604 2011

[10] K C Foucher D E Hurwitz and M A Wimmer ldquoPreop-erative gait adaptations persist one year after surgery inclinically well-functioning total hip replacement patientsrdquoJournal of Biomechanics vol 40 no 15 pp 3432ndash3437 2007

[11] M Baker J Moreside I Wong and D J Rutherford ldquoPassivehip movement measurements related to dynamic motionduring gait in hip osteoarthritisrdquo Journal of OrthopaedicResearch vol 34 no 10 pp 1790ndash1797 2016

[12] M Constantinou A Loureiro C Carty P Mills andR Barrett ldquoHip joint mechanics during walking in individualswith mild-to-moderate hip osteoarthritisrdquo Gait amp Posturevol 53 pp 162ndash167 2017

[13] I Eitzen L Fernandes L Nordsletten and M A RisbergldquoSagittal plane gait characteristics in hip osteoarthritis pa-tients with mild to moderate symptoms compared to healthycontrols a cross-sectional studyrdquo BMC MusculoskeletalDisorders vol 13 no 1 p 258 2012

[14] R J Leigh S T Osis and R Ferber ldquoKinematic gait patternsand their relationship to pain in mild-to-moderate hip os-teoarthritisrdquo Clinical Biomechanics vol 34 pp 12ndash17 2016

[15] I H Reininga M Stevens R Wagenmakers S K BulstraJ W Groothoff and W Zijlstra ldquoSubjects with hip osteo-arthritis show distinctive patterns of trunk movements duringgait-a body-fixed-sensor based analysisrdquo Journal of Neuro-engineering and Rehabilitation vol 9 no 1 p 3 2012

[16] R Needham R Naemi and N Chockalingam ldquoQuantifyinglumbar-pelvis coordination during gait using a modifiedvector coding techniquerdquo Journal of Biomechanics vol 47no 5 pp 1020ndash1026 2014

[17] R Chang R Van Emmerik and J Hamill ldquoQuantifyingrearfoot-forefoot coordination in human walkingrdquo Journal ofBiomechanics vol 41 no 14 pp 3101ndash3105 2008

[18] J F Seay R E A Van Emmerik and J Hamill ldquoInfluence oflow back pain status on pelvis-trunk coordination duringwalking and runningrdquo Spine vol 36 no 16 pp E1070ndashE1079 2011

[19] B C Heiderscheit J Hamill and R E A van EmmerikldquoVariability of stride characteristics and joint coordinationamong individuals with unilateral patellofemoral painrdquoJournal of Applied Biomechanics vol 18 no 2 pp 110ndash1212002

[20] H-J Park T Sim S-W Suh J H Yang H Koo andJ H Mun ldquoAnalysis of coordination between thoracic andpelvic kinematic movements during gait in adolescents withidiopathic scoliosisrdquo European Spine Journal vol 25 no 2pp 385ndash393 2016

[21] J A Smith J M Popovich and K Kulig ldquoe influence of hipstrength on lower-limb pelvis and trunk kinematics andcoordination patterns during walking and hopping in healthywomenrdquo Journal of Orthopaedic amp Sports Physical erapyvol 44 no 7 pp 525ndash531 2014

[22] M A Samaan H-L Teng D Kumar et al ldquoAcetabularcartilage defects cause altered hip and knee joint coordinationvariability during gaitrdquo Clinical Biomechanics vol 30 no 10pp 1202ndash1209 2015

[23] W Kawakami M Takahashi Y Iwamoto and K ShinakodaldquoCoordination among shank rearfoot midfoot and forefootkinematic movement during gait in individuals with halluxvalgusrdquo Journal of Applied Biomechanics vol 35 no 1pp 44ndash51 2019

[24] F Bugane M Benedetti V DrsquoAngeli and A Leardini ldquoEs-timation of pelvis kinematics in level walking based on a singleinertial sensor positioned close to the sacrum validation on


healthy subjects with stereophotogrammetric systemrdquo Bio-medical Engineering Online vol 13 no 1 p 146 2014

[25] D Kobsar S T Osis A Phinyomark J E Boyd andR Ferber ldquoReliability of gait analysis using wearable sensorsin patients with knee osteoarthritisrdquo Journal of Biomechanicsvol 49 no 16 pp 3977ndash3982 2016

[26] I H F Reininga M Stevens R Wagenmakers et alldquoCompensatory trunk movements in patients with hip os-teoarthritisrdquo American Journal of Physical Medicine amp Re-habilitation vol 90 no 8 pp 681ndash687 2011

[27] K Aminian C Trevisan B Najafi et al ldquoEvaluation of anambulatory system for gait analysis in hip osteoarthritis andafter total hip replacementrdquo Gait amp Posture vol 20 no 1pp 102ndash107 2004

[28] S A A N Bolink L R Brunton S van Laarhoven et alldquoFrontal plane pelvic motion during gait captures hip oste-oarthritis related disabilityrdquo Hip International vol 25 no 5pp 413ndash419 2015

[29] J McCamley M Donati E Grimpampi and C Mazza ldquoAnenhanced estimate of initial contact and final contact instantsof time using lower trunk inertial sensor datardquoGait amp Posturevol 36 no 2 pp 316ndash318 2012

[30] H Warashina M Kato S Kitamura T Kusano andY Hasegawa ldquoe progression of osteoarthritis of the hipincreases degenerative lumbar spondylolisthesis and causesthe change of spinopelvic alignmentrdquo Journal of Orthopaedicsvol 16 no 4 pp 275ndash279 2019

[31] T Okuda T Fujita A Kaneuji K Miaki Y Yasuda andT Matsumoto ldquoStage-specific sagittal spinopelvic alignmentchanges in osteoarthritis of the hip secondary to develop-mental hip dysplasiardquo Spine vol 32 no 26 pp E816ndashE8192007

[32] J Schroter V Guth M Overbeck D Rosenbaum andW Winkelmann ldquoe lsquoEntlastungsgarsquong A hip unloadinggait as a new conservative therapy for hip pain in the adultrdquoGait amp Posture vol 9 no 3 pp 151ndash157 1999

[33] K Uemura P R Atkins S A Maas C L Peters andA E Anderson ldquoree-dimensional femoral head coverage inthe standing position represents that measured in vivo duringgaitrdquo Clinical Anatomy vol 31 no 8 pp 1177ndash1183 2018

[34] M Fujii Y Nakashima T Sato M Akiyama and Y IwamotoldquoAcetabular tilt correlates with acetabular version and cov-erage in hip dysplasiardquo Clinical Orthopaedics and RelatedResearch vol 470 no 10 pp 2827ndash2835 2012


dynamic range of motion of the hip during gait had re-portedly improved among post-THR patients it was notcomparable with that in normal individuals [10] Otherstudies also examined the gait of patients with hip OA whohad not undergone THR by studying joint kinematics ki-netics and their correlations [4 11ndash15] But it remainsunclear whether the hip joint stiffness is present in the gait ofpatients with hip OA preoperatively Traditional measure-ments such as examining the range of motion during gaitare not enough for understanding control mechanisms It isimportant to examine the gait and movement patterns ofpatients with hip OA by focusing on time interval such asjoint coordination [16] e vector coding technique (VCT)allows researchers to examine the coordination pattern ofthe movement over time [17 18] Several studies found thatthe people with pain or musculoskeletal disorders such aspatellofemoral pain low back pain or idiopathic scoliosisshowed less movement between two relevant segments orjoints using the VCT [18ndash20] Smith et al [21] investigatedthe coordination pattern between pelvis and thigh with weakhip muscle in healthy subjects and Samaan [22] studied thecoordination pattern between hip and knee movement inhip OA patients with VCTtechnique However there was nostudy that investigated the thighndashpelvisndashlumbar coordina-tion pattern of patients with hip OA preoperatively eVCT characterizes the movement coordination pattern oftwo segments or joints as the predominance of movementspeed using the following four phases [16 23] in-phase (thesame directional movement with similar movement speed)anti-phase (opposite directional movement with similarmovement speed) proximal-phase (relatively large proximaljointsegment movement speed) and distal-phase (relativelylarge distal jointsegment movement speed) e ldquoin-phaserdquomovement especially indicates a particular movementpattern with similar movement speed between two segmentsor joint [16 23] e ldquoin-phaserdquo and ldquostiffrdquo are not actuallyequal in point of considering the force applied former re-ferring to the observed movement only and latter referringto the movement as the response to the force appliedHowever it has been conceivable that the in-phase coor-dination pattern relates to the stiffened condition [20] Webelieve that using the VCT to directly assess the patterns ofmovement for patients with hip OA will give us theknowledge regarding the stiff hip movement during thepelvicthigh ldquoin-phaserdquo coordination pattern as well as therelationship between the hip and relevant region movementsover time in patients with hip OA

We measured changes in segmental movement using aninertial measurement unit (IMU) e IMU is inexpensiveand easy to use in a clinical setting because the IMU does notneed the laboratory setting which is needed for using motioncapture system Other groups have documented the reli-ability and validity of using an IMU for measuring gaitkinematics [24ndash26] and assessing the symptoms of mus-culoskeletal disorders [15 25 27 28] In the first step ofVCT the change in the angles of two segmentaljointmovements can be calculated e procedure thus enablesthe calculation of the angular velocity which is the differencebetween two successive angles us we decided to use this

procedure in addition to collecting angular velocity datausing IMU

e aim of this study was to investigate the coordinationpattern of segmental movements in the gait of patients withhip OA using VCTand IMU which allows us to suggest thepresence of stiffness and the relationship of each segmentalmovement We hypothesized that patients with hip OA willreveal increased rate of ldquoin-phaserdquo movement pattern due tostiffness in the hip joint We also hypothesized that the in-phase coordination pattern of pelvicthigh movementswould have an anteriorposterior tilt due to limited hipextension movements among patients with hip OA

2 Materials and Methods

21 Participants e present study included 11 patientswith hip OA from an outpatient clinic and 11 age- andgender-matched community-dwelling control volunteerse inclusion criteria of the hip OA group were as followsbeing female ability to walk without the aid of medicalequipment age 40ndash70 years and having a grade of 3 or 4 onthe Kellgren-Lawrence grading system ere were very fewmale patients and the presence of differences with respect topatient sex was anticipated therefore we did not includemale patients in the study Patients with neurologicalconditions and other lower extremity joint disorders whichaffected the required tasks were excluded from the studyAmong patients with bilateral hip OA the more affected sidewas chosen as the ldquotargetedrdquo side e control group in-cluded volunteers who did not have any complaint abouttheir hips e other criteria were the same as those for hipOA group Data on study participants are found in Table 1

All procedures were approved by the institutional ethicscommittee and all participants provided written informedconsent prior to participating in this studye experimentalprocedures of the study were conducted according to theDeclaration of Helsinki

22 Data Collection ree IMUs (TSND151 ATR-pro-motions Soraku Japan) each equipped with a triaxial ac-celerometer gyroscope andmagnetometer were attached tothe following areas on the lateral side of the thigh betweenthe femoral condyle and greater trochanter (thigh) on thedorsal side between both posterior superior iliac spines(pelvis) and on the dorsal side at the first lumbar spinousprocess (lumbar) Sensor axes were aligned to carefullydetect the anterior-posterior medial-lateral and superior-inferior directions of movements e sampling rate was setat 100Hz

e participants were required to perform a uniform setof tasks First the participants were asked to stand quietlyAfter standing quietly they were directed to begin walkingover a distance of 30m following an oral cue by the ex-aminer Participants were asked to walk the path until theexaminer gave them an oral cue to stop Once directed tostop walking participants were required to maintain a quietstanding posture until given an oral cue by the examinerParticipants did not wear shoes during these tasks and were






ci tanminus1 αi

βi

1113888 1113889 (1)



xi 1n

1113944

n

i1cos ci( 1113857 (2)

yi 1n

1113944

n

i1sin ci( 1113857 (3)

ci

arctanyi

xi

1113888 1113889 xi gt 0

180 + arctanyi

xi

1113888 1113889 xi gt 0

⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(4)


δi

xi2

+ yi2

1113969

(5)

CAVi

2 middot 1 minus δi1113872 1113873

1113969

middot180π

(6)








3 Results






4 Discussion



Distal-phase

In-phase

In-phase




1125 le γ lt 1575 2925 le γ lt 3375

Anti-phase

Anti-phase


(degs)



angle

Distal-phase

Proximal-phase

225

6751125

1575

2025

2475 2925

3375





80

70

60

50

40

30

20

10

0

Rate

()

lowast

lowast


ControlHip OA

(a)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(b)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(c)

70

60

50

40

30

20

10

0

Rate

() lowast

lowast


ControlHip OA

(d)

100

80

60

40

20

0

Rate

()

lowast


ControlHip OA

(e)

70

60

50

40

30

20

10

0

Rate

()

lowast


ControlHip OA

(f )



3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )










Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References










































ci tanminus1 αi

βi

1113888 1113889 (1)



xi 1n

1113944

n

i1cos ci( 1113857 (2)

yi 1n

1113944

n

i1sin ci( 1113857 (3)

ci

arctanyi

xi

1113888 1113889 xi gt 0

180 + arctanyi

xi

1113888 1113889 xi gt 0

⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(4)


δi

xi2

+ yi2

1113969

(5)

CAVi

2 middot 1 minus δi1113872 1113873

1113969

middot180π

(6)








3 Results






4 Discussion



Distal-phase

In-phase

In-phase




1125 le γ lt 1575 2925 le γ lt 3375

Anti-phase

Anti-phase


(degs)



angle

Distal-phase

Proximal-phase

225

6751125

1575

2025

2475 2925

3375





80

70

60

50

40

30

20

10

0

Rate

()

lowast

lowast


ControlHip OA

(a)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(b)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(c)

70

60

50

40

30

20

10

0

Rate

() lowast

lowast


ControlHip OA

(d)

100

80

60

40

20

0

Rate

()

lowast


ControlHip OA

(e)

70

60

50

40

30

20

10

0

Rate

()

lowast


ControlHip OA

(f )



3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )










Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References







































3 Results






4 Discussion



Distal-phase

In-phase

In-phase




1125 le γ lt 1575 2925 le γ lt 3375

Anti-phase

Anti-phase


(degs)



angle

Distal-phase

Proximal-phase

225

6751125

1575

2025

2475 2925

3375





80

70

60

50

40

30

20

10

0

Rate

()

lowast

lowast


ControlHip OA

(a)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(b)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(c)

70

60

50

40

30

20

10

0

Rate

() lowast

lowast


ControlHip OA

(d)

100

80

60

40

20

0

Rate

()

lowast


ControlHip OA

(e)

70

60

50

40

30

20

10

0

Rate

()

lowast


ControlHip OA

(f )



3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )










Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References








































80

70

60

50

40

30

20

10

0

Rate

()

lowast

lowast


ControlHip OA

(a)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(b)

80

70

60

50

40

30

20

10

0

Rate

()


ControlHip OA

(c)

70

60

50

40

30

20

10

0

Rate

() lowast

lowast


ControlHip OA

(d)

100

80

60

40

20

0

Rate

()

lowast


ControlHip OA

(e)

70

60

50

40

30

20

10

0

Rate

()

lowast


ControlHip OA

(f )



3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )










Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References






































3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(a)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(b)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(c)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(d)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(e)

3603375

2925

2475

2025

1575

1125

675

2250

CA an

d CA

V (d

eg)

0 20 40 60 80 100Time ()




(f )










Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References













































Lumbar



Pelvis



igh






5 Conclusion


Data Availability




References







































5 Conclusion


Data Availability




References






































coordinationpatternofthethigh,pelvic,andlumbar

Documents