an investigation of the relationship between hindlimb ... & dyson saddle slip... · an...

An investigation of the relationship between hindlimb lameness andsaddle slipL. GREVE and S. J. DYSON*

Centre for Equine Studies, Animal Health Trust, Newmarket, Suffolk, UK.

*Correspondence email: [email protected]; Received: 17.08.12; Accepted: 07.12.12

Summary

Reasons for performing study: We have observed saddle slip consistently to one side because of a crooked rider, an ill-fitting saddle, asymmetry in ahorse’s thoracolumbar shape and lameness. Currently, there are no objective data to permit assessment of the relative importance of each factor.Objectives: To document the frequency of occurrence of saddle slip in horses with hindlimb lameness compared with other horses. To describe the effect oflameness characteristics and grade, the abolition of lameness by diagnostic analgesia, breed, size, thoracolumbar shape and symmetry and the rider’sweight.Methods: One hundred and twenty-eight horses were assessed prospectively, and lameness and saddle slip were assigned a grade before and afterdiagnostic analgesia. The thoracolumbar shape and symmetry were measured objectively. In 3 horses, the force distribution and magnitude underneath thesaddle were measured before and after diagnostic analgesia.Results: The saddle consistently slipped to one side in 38 of 71 horses (54%) with hindlimb lameness, compared with one of 26 horses (4%) with forelimblameness, none of 20 (0%) with back pain and/or sacroiliac joint region pain and none of 11 sound horses (0%). The association between saddle slip andhindlimb lameness was significant (Spearman’s rank correlation coefficient, r = 0.548, P<0.001). Diagnostic analgesia abolishing the hindlimb lamenesseliminated the saddle slip in 37 of 38 horses (97%). In 2 horses, the saddle continued to slip after resolution of lameness; one horse had bilateral forelimblameness and the other horse had concurrent hindlimb and forelimb lameness. The saddle of both these horses was asymmetrically flocked. The saddleslipped to the side of the lamer hindlimb in most horses (32 of 37 [86%]). No horse with saddle slip had significant left–right asymmetry of the back at 4predetermined sites.Conclusions and clinical relevance: Hindlimb lameness is an important factor in inducing saddle slip. Saddle slip may be an indicator of the presence ofhindlimb lameness.

Keywords: horse; saddle; back movement; pressure measurements

Introduction

Saddle slip in sports horses is a well-recognised problem that can occurfor a variety of reasons, including asymmetry in the shape of the horse’sback [1], riders sitting crookedly [2–4], ill-fitting saddles [5,6] and hindlimblameness [7,8]. A pilot study revealed altered force distribution underneaththe saddle in horses with hindlimb lameness and back pain [9]. Symmetricalthoracolumbar movement in sound horses has been demonstrated usingkinematic analysis [10,11]. However, symmetry and amplitude of themovement of the back are influenced either by experimentally inducedtransient forelimb or hindlimb lameness [12–15] or by experimentallyinduced back pain [16,17]. These studies did not induce concurrent backpain and lameness, although their coexistence is commonly recognisedclinically [18]. The majority of published information on thoracolumbarkinematics is from treadmill studies [11,16,17], and there are limited dataon ridden sound [19–22] and lame horses [9] in ‘real-life’ conditions.

It has been observed in ridden horses that in association with hindlimblameness saddles may slip consistently to one side. This has encouragedowners to try multiple saddles and additional padding to try to diminish thesaddle slip, but this has generally been ineffective [7]. Saddle slip may beinfluenced by the rider’s weight and the back shape of the horse, withgreater slip occurring with lighter weight riders and in native-type horseswith round back contours [7].

A comprehensive description of the saddle kinematics in ridden soundand lame horses outside gait laboratories has not been reported. Althoughinterface force–distribution data have confirmed that there are a variety offactors that influence the movement of a saddle [6,20,23,24], there islimited knowledge of the interrelationships between the horse, saddle andrider and no objective assessment of saddle slip and gait irregularities inlame horses. By evaluating the response to diagnostic analgesia, we areable objectively to separate out the relative importance of the movement ofthe horse when ridden and other factors that potentially may induce saddleslip.

The aims of the present study were as follows: 1) to document thefrequency of occurrence of saddle slip in horses with hindlimb lameness,

compared with horses without hindlimb lameness; 2) to describe the effectof lameness grade, unilateral or bilateral hindlimb lameness, multilimblameness (�3 lame limbs) and the coexistence of sacroiliac joint regionpain and/or thoracolumbar-sacral pain/stiffness; 3) to document the effectof abolition of lameness by diagnostic analgesia; and 4) to determine therelationships between saddle slip and the horse’s thoracolumbar shapeand evaluate the influence of rider weight, horse breed, type, bodycondition score (BCS) and size.

It was hypothesised that saddle slip may be induced by hindlimblameness, with the saddle slipping most frequently to the side of the lame(r)limb, and that the degree of saddle slip may be reduced by improvement inthe lameness by diagnostic analgesia.

Materials and methods

A prospective study was performed at the Animal Health Trust, approvedby the Ethics Committee and with informed consent of owners, betweenAugust 2011 and August 2012. All horses that were presented forgait assessment and were ridden by at least 2 riders were included.Diagnosis was assigned prospectively based on the results of acomprehensive clinical evaluation, diagnostic analgesia and imaging.The horses were divided into the following 5 categories: Group A, hindlimblameness; Group B, forelimb lameness; Group C, horses with both forelimband hindlimb lameness; Group D, poor performance in associationwith thoracolumbar-sacral and/or sacroiliac joint region pain; and Group E,sound horses.

Horse dataAge, breed, sex, height, bodyweight, BCS [25], work discipline and thepresence of hindquarter and epaxial muscle atrophy were recorded. Aninterobserver repeatability study was carried out on 10 horses for BCS,to confirm that the method could be used practically. In addition, anintraobserver repeatability study was performed, with 5 horses assessed3 times in random order. A variance of less than 1% was obtained.

bs_bs_banner

Equine Veterinary Journal ISSN 0425-1644DOI: 10.1111/evj.12029

1Equine Veterinary Journal •• (2013) ••–•• © 2012 EVJ Ltd

Assessment of thoracolumbar shape and symmetryThe thoracolumbar shape and symmetry were measured at the level of the18th (T18), the 13th (T13) and the 8th (T8) thoracic vertebrae, identified bypalpation of the ribs, and at a site one-third of the distance between thepoint of the elbow and the point of the shoulder (shoulder region). AFlexible Curve Rulera was shaped around the dorsum, perpendicular tothe dorsal midline, to follow the body contours. The resultant shape wasdrawn on graph paper. A single investigator (L.G.) performed all themeasurements. Repeatability of the measuring technique was tested witha pilot repeatability study on 3 differently shaped and sized horsesmeasured 10 times in random order, followed by 5 measurements inrandom order of an additional 10 horses (measurement error �2 mm). Theminimal detectable difference outside of the measurement error wasestablished using coefficients of variation (CV) of 4, 8 and 12% (95%confidence interval [CI]) across all horses [26], and the number of horseswith asymmetries greater than a CV of 4, 8 and 12% (95% CI) wasdetermined. Asymmetries between left and right sides were confirmed byrotating the Flexible Curve Ruler by 180° in a horizontal plane andcomparing it with the predrawn curve. All measurements were performedwith the horses standing squarely on a level surface, before exercise.Preliminary data had indicated that alterations in stance could altermeasurements slightly and that measurements acquired soon afterexercise varied compared with before exercise.

Thoracolumbar shape categoriesThe thoracolumbar shape at each measurement site was divided into thefollowing 4 categories: type 1, concave; type 2, straight; type 3, convex;and type 4, very convex. The grading system was based on the ratiobetween the horizontal distances between left and right sides 3 and 15 cm

ventral to the dorsal midline (Fig 1). Type 1 back shape was defined ashaving a ratio �0.30 (Fig 1a), type 2 back shape a ratio of 0.31 � 0.35(Fig 1b), type 3 back shape a ratio of 0.36 � 0.40 (Fig 1c) and type 4 backshape a ratio of �0.41 (Fig 1d).

Description of lameness characteristics andsaddle slipAll horses were examined by one experienced clinician (S.J.D.) moving inhand, on the lunge and ridden, and the lameness was graded in eachcircumstance (0 = sound, 2 = mild, 4 = moderate, 6 = severe and 8 =nonweightbearing) [27]. The horses were evaluated when ridden by theowner and at least one experienced rider from the Animal Health Trust. Thelameness was graded at sitting and rising trot on both left and rightdiagonals, on the left and right reins, and when performing specificmovements, such as 10 m diameter circles to the left and right. Thelameness was graded before and after diagnostic analgesia for eachcircumstance [27]. Diagnostic analgesia was performed in all lame limbs.Rider straightness, rider weight, the grade of saddle slip, whether itoccurred with more than one rider, and whether saddle slip was influencedby the direction of movement or by the diagonal on which the rider wassitting were recorded. Both authors assessed saddle slip by consensus.The same rider rode the horse before and after diagnostic analgesia. Thegrade of saddle slip after lameness had been abolished was recorded.Photographs and video recordings were obtained.

A pilot study to determine accuracy of detection of saddle slip hadpreviously been performed, which had involved placing markers on horses,saddles and riders. All horses were videoed being ridden in a straight line.The presence or absence of saddle slip was determined without and withmarkers by 3 people independently. There was complete agreement

a) c)

b) d)

The

dis

tanc

e ve

ntra

l to

the

dor

sal m

idlin

e (c

m)

The

dis

tanc

e ve

ntra

l to

the

dor

sal m

idlin

e (c

m)

The

dis

tanc

e ve

ntra

l to

the

dor

sal m

idlin

e (c

m)

-21-23 -19 -17 -15 -13 -11 -9 -7 -5 -3 -1

-1-2-3-4-5-6-7-8-9

-10-11-12-13-14-15-16-17 (cm)

Back shape 1concave

(0.21–0.30)

Back shape 2straight

(0.31-0.35)

Back shape 4very convex(0.41–0.49)

0

-1-2-3-4-5-6-7-8-9

-10-11-12-13-14-15-16-17 (cm)

0

1 3 5 7 9 11 13 15 17 19 21 23 -23-25-27

The

dis

tanc

e ve

ntra

l to

the

dor

sal m

idlin

e (c

m)

-21-19-17-15-13-11-9 -7 -5 -3 -1

-1-2-3-4-5-6-7-8-9

-10-11-12-13-14-15-16-17 (cm)

Back shape 3convex

(0.36–0.40)

0

-1-2-3-4-5-6-7-8-9

-10-11-12-13-14-15-16-17 (cm)

0

1 3 5 7 9 11 13 15 17 19 21 23 25 27

-23-25-27 -21-19-17-15-13-11-9 -7 -5 -3 -1 1 3 5 7 9 11 13 15 17 19 21 23 25 27-23-25-27 -21-19-17-15-13-11-9 -7 -5 -3 -1 1 3 5 7 9 11 13 15 17 19 21 23 25 27

Fig 1: Four categories of the thoracolumbar shape. The grading system is based on the ratio between the horizontal distances between left and right sides 3 and 15 cm ventral to thedorsal midline. The black and grey lines represent the range of ratios within each shape type. Type 1, concave, defined as having a ratio �0.30 (a); type 2, straight, with a ratio of 0.31� 0.35 (b); type 3, convex, with a ratio of 0.36 � 0.40 (c) and type 4, very convex, with a ratio �0.41 (d).

Saddle slip and lameness L. Greve and S. J. Dyson

2 Equine Veterinary Journal •• (2013) ••–•• © 2012 EVJ Ltd

between the 3 observers, and for observations made with or withoutmarkers.

An overall saddle slip grade was obtained for each horse using a scalefrom 0 to 2: grade 0, no saddle slip at any gaits; grade 1, saddle slip at thetrot and canter on one or both reins (Fig 2); and grade 2, obvious saddle slipat trot and canter on one or both reins, such that the rider intermittentlyhad to stop the horse and straighten the saddle. The saddle slip grade wasapplied to the lightest weight rider for all horses. The saddle slip was alsoassessed subjectively in different circumstances (e.g. straight lines andcircles, and with riders of different bodyweights). In 3 horses, the forcedistribution and magnitude underneath the saddle was measured using aCONFORMatb pressure mat [28].

Assessment of the saddle fitThe tree of the saddle was checked for evenness and symmetry. Theleft–right symmetry of the panels and presence of lumps or depressionswere recorded [1]. The saddle area was palpated carefully for any signs offocal muscle soreness. An experienced clinician evaluated the saddle fitvisually and following the ridden examination. Focal areas of reducedsweating in the saddle area was recorded [29]. The saddle type (tree,flexible or rigid, �gullet) and any special flocking, pads and numnahs usedwere recorded.

Data analysisThe proportion of horses with saddle slip and hindlimb lameness, forelimblameness and a combination was determined. Saddle slip was comparedamong Groups A, B, C, D and E using Cochran’s Q test, a chi-squared test orFisher’s exact test, as appropriate.

For all horses in the study, Spearman’s rank correlation was used to testfor association between saddle slip and the following factors: 1) unilateralhindlimb lameness; 2) bilateral hindlimb lameness; 3) concurrent unilateralhindlimb and forelimb lameness; 4) 3-limb lameness; 5) 4-limb lameness;6) unilateral or bilateral hindlimb lameness and concurrent sacroiliacjoint region pain, verified by periarticular diagnostic analgesia [30];7) concurrent hindlimb lameness and the presence of osseous pathology

in the thoracolumbar region; 8) concurrent hindlimb lameness andclose/impinging spinous processes (improved gait was seen after localanalgesia of the spinous processes); 9) unilateral gluteal muscle atrophyand/or epaxial muscle atrophy; and 10) tension in the epaxial musculature.

For horses in Groups A and C, Spearman’s rank correlation was used totest for association between the grade of lameness and the followingfactors: 1) grade of saddle slip; 2) the tendency of the saddle to slip towardsthe sound or less lame hindlimb; 3) sex; 4) age; 5) breed; 6) work discipline;7) BCS; and 8) weight/height ratio.

For all horses in Groups A and C, Spearman’s rank correlation was used totest for association between saddle slip and the following factors: 1) theback shape at the shoulder, T8, T13 and T18; and 2) the effect of having asimilar or wider back shape at the level of T13 compared with T18 (horseswere selected for inclusion if the back contour at T13 � T18). All statisticalanalyses were made using SPSS Statistics 20c, with significance set atP<0.05.

Results

There were 11 sound horses and 117 with lameness or poor performance.The horses ranged in age from 5 to 17 years (mean 8.72 years, median 8years) and comprised 92 geldings and 36 mares. The duration of lamenessranged from 0 to 12 months (mean 1 month, median 1.5 months). Horseswere used for dressage (n = 32), showjumping (n = 26), eventing (n = 29),general purposes (n = 40; horses used for unaffiliated competitions wereclassified as general purpose) and endurance (n = 1). Breeds comprisedWarmbloods (n = 56), Thoroughbreds (n = 9), Thoroughbred crosses(n = 28), cobs (n = 6), ponies (n = 15) and others (n = 14).

The horses ranged in weight from 286 to 914 kg (mean 568 kg, median574 kg), in height from 117 to 183 cm (mean 167 cm, median 170 cm) andthe weight/height ratio ranged from 1.9 to 4.8 kg/cm (mean 3.4 kg/cm,median 3.4 kg/cm). The BCS was 1 (n = 1), 2 (n = 10), 3 (n = 106) and 4(n = 11). No horses with a well-fitted saddle had focal areas of reducedsweating in the saddle region. Thirty-nine horses had saddle slip,comprising 34 horses with grade 1 and 5 with grade 2, including 2 inducedby an asymmetric saddle. Thirteen of 39 riders (33%) had appreciated thepresence of saddle slip prior to the clinical investigation. The lamenessgroups are summarised in Figure 3, together with the number of horseswith saddle slip for each lameness group. Twenty horses had unilateralhindlimb lameness and 51 horses had bilateral hindlimb lameness. Thehindlimb lameness grade when ridden ranged from 1 to 6; the mostfrequent lameness grade was 2 (n = 33). In 3 horses, hindlimb lamenessgrade was at least one grade different when the rider changed from sittingon one diagonal to the other, irrespective of whether the horse was on theleft or right rein, and this influenced the tendency for the saddle to slip.

The weight of the normal rider of horses with saddle slip associated withhindlimb lameness ranged from 47 to 80 kg (mean 63 kg, median 65 kg).Five riders sat crookedly. The weight of the riders (n = 4) from the AnimalHealth Trust ranged from 43 to 60 kg (mean 56 kg, median 60 kg).

Saddle slipThere was complete consensus between the authors in assessment ofsaddle slip. Saddle slip occurred with at least 2 riders in all the horses withsaddle slip. In horses with saddle slip which was abolished by resolution ofhindlimb lameness, saddle slip was consistently greatest with a lighterweight rider (n = 37); however, when saddle slip was related to anasymmetrical saddle, slip was greater with a heavier weight rider (n = 2). InGroup A, the saddle consistently slipped to one side in 20 of 36 horses(56%); 18 of 35 horses (51%) had saddle slip in Group C, compared with oneof 26 (4%) in Group B, none of 20 (0%) in Group D and none of 11 (0%) inGroup E. The association between saddle slip and hindlimb lameness wassignificant (Spearman’s r = 0.548, P<0.001). The association betweensaddle slip and horses with hindlimb lameness alone (r = 0.354, P<0.001)and horses with hindlimb lameness and coexistent forelimb lameness(r = 0.254, P = 0.004) was significant.

Objective relationship between saddle slipand lamenessThere was a significant difference in the occurrence of saddle slip amongGroups A, B, C and D (P<0.001; Fig 3). For all horses, there was a significant

a) b)

Fig 2: Caudal images of a horse ridden in straight lines on the left rein (a) and on a leftcircle (b). There is saddle slip to the right, which was worse in circles compared withstraight lines.

L. Greve and S. J. Dyson Saddle slip and lameness


positive association between the presence of saddle slip and bilateralhindlimb lameness (r = 0.344, P<0.001), unilateral hindlimb lameness (r =0.286, P = 0.001), concurrent unilateral hindlimb and unilateral forelimblameness (r = 0.222, P = 0.012) and concurrent bilateral hindlimb lamenessand sacroiliac joint region pain (r = 0.220, P = 0.013). The associationbetween saddle slip and 3-limb lameness (r = 0.116, P = 0.191), 4-limblameness (r = 0.098, P = 0.269) and coexistent unilateral hindlimb lamenessand sacroiliac joint region pain (r = 0.012, P = 0.890) was nonsignificant.There was a significant association between the presence of saddle slip andtense epaxial muscles in the thoracolumbar region (r = 0.358, P<0.001) andbetween unilateral hindquarter muscle atrophy and/or epaxial muscleatrophy (r = 0.274, P = 0.002). There were significant associations betweenthe presence of saddle slip and both hindlimb lameness and concurrentosseous pathology in the thoracolumbar region (r = 0.436, P<0.001) andhindlimb lameness and improvement in clinical signs after local analgesia ofclose/impinging spinous processes (r = 0.341, P<0.001). Horses withunilateral hindquarter muscle atrophy and/or epaxial muscle atrophy (r =0.472, P<0.001) and horses with sacroiliac joint region pain (r = 0.290, P =0.001) were more likely to have tense epaxial muscles in the thoracolumbarregion compared with horses in Group B (r = -0.250, P = 0.005), in which theassociation was negative.

For all horses in Groups A and C, there was no significant associationbetween the grade of saddle slip and lameness grade (r = -0.179, P = 0.136),nor was there a significant association between the lameness grade and thetendency of the saddle to slip towards the sound or less lame hindlimb(r = -0.136, P = 0.260). The causes of hindlimb lameness are summarised inTable 1, together with the number of horses with saddle slip for each cause.

Objective relationship between saddle slip andhorse dataThere was no significant association between saddle slip and breed (r =-0.069 P = 0.565), BCS (r = 0.069, P = 0.570), weight/height ratio (r = 0.151,P = 0.210), work discipline (r = -0.064, P = 0.596), sex (r = -0.150, P = 0.211)or age (r = 0.087, P = 0.472).

Association between back shape and saddle slipSubjectively, only minor asymmetries <2 cm between left and right sideswere observed by rotating the Flexible Curve Ruler through 180° in a

horizontal plane and comparing it over the predrawn curve. Greaterasymmetries were observed in horses without saddle slip than in horseswith saddle slip.

The mean typical measurement error for each level (shoulder, T8, T13and T18) across all horses included in the pilot repeatability studyresulting in minimal detectable difference in back shape between left andright sides was 0.6 cm for a CV of 4%, 1.2 cm for a CV of 8% and 1.8 cm fora CV of 12%. All the horses with saddle slip had asymmetries between theleft and right sides of less than a CV of 8% (1.2 cm) measured at the levelof the shoulder region, T8, T13 and T18. Five horses had asymmetriesgreater than a CV of 8% (1.2 cm) in the saddle region, and none hadsaddle slip.

There was no significant association between saddle slip and the type ofback shape at the shoulder, T8, T13 or T18 (r = 0.128, P = 0.289), nor wasthere a significant association between having a similar type of back shapeat T8, T13 and T18 and saddle slip (r = 0.112, P = 0.352) or having one or lessdifference in the type of back shape among T8, T13 and T18 (r = 0.150, P =0.211). However, there was a significant positive association betweensaddle slip and horses with a wide back shape at T13 (defined as having aback contour at T13 � T18; r = 0.340, P = 0.004). There was also asignificant positive association between horses having one or lessdifference in the type of back shape among T8, T13 and T18 and a wide backshape at T13 (r = 0.288, P = 0.015).

Factors affecting saddle slip associated withhindlimb lamenessThirty-seven horses had saddle slip caused by hindlimb lameness. Thesaddle slipped to the side of the lame or lamer hindlimb when the horse wasworked in straight lines, going large around the arena and also in circles of20 m diameter in most horses (32 of 37 [86%]). The direction of the saddleslip was associated with the rein on which the horse appeared most lame in18 of 37 horses (49%). However, this was complicated by the fact that someof the horses (13 of 37 [35%]) going large or in a 20 m circle appeared lamerwith the lame limb on the outside, whereas on a 10 m circle the horseappeared lamer with the lame limb on the inside. However, the difference inappearance of the lameness did not alter the direction of saddle slip. Saddleslip was present on both reins in 6 of 37 horses (16%) and on only one reinin 31 of 37 (84%).

26

24

22

20

18

16

14

12

10Num

ber

of h

orse

s

8

6

4

2

0

26

24

22

20

18

16

14

12

10

8

6

4

2

0Group A:

HL Iamenesswithout

coexistent FLlameness

Group B:FL Iameness

withoutcoexistent HL

lameness

Group C:Concurrent FL

and HLlameness

Group D:Back and/or SI

region pain

Group E:Sound horses

Fig 3: The frequency of saddle slip within the 5categories of horses. Open columns represent horseswithout saddle slip and filled columns horses withsaddle slip consistently to one side before lamenesswas abolished by diagnostic analgesia.Abbreviations: FL, forelimb; HL, hindlimb;and SI, sacroiliac joint.



Saddle slip was consistently greater in circles compared with straightlines (37 of 37 horses [100%]), irrespective of the appearance of thelameness (Fig 2). Lameness was worse when the rider sat on the diagonal ofthe lame(r) limb in the stance phase in 11 of 37 horses (30%), but this wasassociated with greater saddle slip in only 3 horses. Saddle slip wasconsistently greater in rising trot compared with sitting trot (37 of 37) andwas remarkably reduced when riding without stirrups (4 of 4). In canter, thesaddle slip was similar to or greater than in rising trot. The saddle slip was

consistently less with a heavier rider in 37 horses ridden by 2 riders (Rider 1= 43 kg and Rider 2 �60 kg), resulting in >17 kg difference in bodyweight,but the grade (0–2) of saddle slip did not change between riders.

Diagnostic analgesia and ill-fitting saddlesDiagnostic analgesia abolishing the lameness did not eliminate the saddleslip in 2 horses, one with hindlimb lameness and one with forelimb

TABLE 1: The causes of hindlimb lameness and prevalence of saddle slip in 71 horses with hindlimb lameness that were ridden by at least 2 people

The causes of hindlimb lameness and other causes of lameness or poor performance

Number ofhorses withthe injury

Number ofhorses withsaddle slip

Percentagewithsaddle slip

Bilateral PSD and FL lameness 13 5 38

Bilateral PSD and SI pain 8 4 50

Bilateral PSD 7 2 29

Unilateral CD and TMT OA 4 2 50

Unilateral FT OA and FL lameness 2 2 100

Bilateral PSD and impinging/overriding 14th–15th to 16th–17th thoracic SPs (one improved withinfiltration of local anaesthetic solution around the impinging/overriding SPs; one was not blocked)� OA 15th–16th and 16th–17th thoracic APJs

2 2 100

Bilateral PSD and SI pain and impinging 10th–15th thoracic SPs (one improved with infiltration of localanaesthetic solution around the impinging SPs; one was not blocked) � OA 15th–16th and16th–17th thoracic APJs

2 2 100

Bilateral PSD and SI pain and FL lameness 6 1 17

Unilateral PSD and FL lameness 4 1 25

Unilateral TMT/CD OA and bilateral PSD 2 1 50

Unilateral TMT/CD OA and FL lameness 1 1 100

Bilateral TMT/CD OA 1 1 100

Unilateral FT OA and middle patellar desmitis 1 1 100

Unilateral FT OA and bilateral PSD 1 1 100

Unilateral CD OA 1 1 100

Bilateral TMT/CD OA and OA 15th–16th thoracic APJs 1 1 100

Unilateral talocalcaneal joint OA, DDFT injury and old fracture of right tuber coxae and FL lameness 1 1 100

Bilateral MTP OA and unilateral lateral oblique sesamoidean desmitis and FL lameness 1 1 100

Lateral suspensory branch desmitis, apical sesamoid fracture of lateral PSB and bilateral PSD; closeSPs between the 13th and 18th thoracic vertebrae (improved with infiltration of local anaestheticsolution around the close SPs) and OA 16th-17th and 17th-18th thoracic APJs; old fracture of thecaudoventral aspect of the 4th cervical vertebra

1 1 100

Unilateral medial collateral desmitis of DIP, unilateral PSD and FL pain 1 1 100

Unilateral accessory desmitis of SL, bilateral PSD and FL lameness 1 1 100

Unilateral accessory desmitis of SL, bilateral PSD and FL lameness; close/impinging SPs of first to 3rdlumbar vertebrae*

1 1 100

Bilateral PSD, SI pain and impinging 10th–15th thoracic SPs (improved with infiltration of localanaesthetic solution around the close SPs) and OA 15th–16th and 16th–17th thoracic APJs and OAcaudal cervical APJs

1 1 100

Unilateral PSD, SI pain, impinging SPs of the 13th thoracic to 2nd lumbar vertebrae (improved withinfiltration of local anaesthetic solution around the impinging SPs) and OA 15th-16th and16th-17th thoracic APJs

1 1 100

Bilateral PSD, SI pain and impinging SPs of the 12th thoracic to 3rd lumbar vertebrae (improved withinfiltration of local anaesthetic solution around the impinging SPs) and FL lameness

1 1 100

Unilateral PSD 2 0 0

Unilateral FT OA and unilateral PSD 1 0 0

Unilateral lateral collateral desmitis of MTP 1 0 0

Unilateral displaced fractures of the distal aspect of the navicular bone and OA caudal cervical APJsand FL lameness

1 0 0

Bilateral PSD, SI pain and asymmetric OA of APJs between the 15th–16th and 16th–17th thoracicvertebrae, supraspinous ligament injury from 15th to 17th thoracic vertebrae* and FL lameness

1 0 0

Total 71 37 52

Abbreviations: APJ(s) = articular process joint(s); CD = centrodistal joint; DDFT = deep digital flexor tendon; DIP = distal interphalangeal joint; FL = forelimb; FT =femorotibial joint; MTP = metatarsophalangeal joint; OA = osteoarthritis; PSB = proximal sesamoid bone; PSD = proximal suspensory desmopathy; SI = sacroiliacjoint region; SL = suspensory ligament; SP(s) = spinous process(es); and TMT = tarsometatarsal joint. *Infiltration of local anaesthetic solution around theclose/impinging SPs was not performed.



lameness; both had an ill-fitting saddle. When ridden with correctly fittingsaddles, no saddle slip was apparent.

Force distribution measurementsTwo horses with lameness and saddle slip that were assessed with aCONFORMatb pressure mat exhibited an asymmetric pressure distributionunderneath the saddle. The mean pressure in the quadrant of the lamerhindlimb during the lamer diagonal stance phase was significantly lowercompared with the corresponding pressure in the quadrant of the sound orless lame hindlimb during the sound or less lame diagonal stance phase.Another horse with bilateral hindlimb lameness exhibited an almostsymmetric pressure distribution underneath the saddle; there was nosaddle slip.

When the horses with saddle slip were re-evaluated following diagnosticanalgesia that abolished the lameness, the difference in pressureunderneath the saddle between lame and sound diagonal stancedisappeared, the amplitude of the sinusoidal curve increased and thecraniocaudal excursion of the centre of pressure (COP) increased, whereasthe right–left excursion and sideways movement of the saddle decreased.All findings from the pressure measurements correlated with the clinicalassessments.

Discussion

In accordance with our hypotheses, saddle slip was induced by hindlimblameness in 37 of 71 horses (52%). The saddle slip was eliminated when thehindlimb lameness was abolished by diagnostic analgesia in 37 of 38 (97%),verifying a causal relationship. Although no association was found betweenthe severity of lameness and saddle slip, horses with severe lamenesswhen exercised in hand were not evaluated ridden and were therefore notincluded in the study. To an untrained observer, some of the horses withobvious saddle slip when ridden appeared clinically normal, and only askilled, experienced clinician recognised subtle lameness. Riders werecommonly surprised by the complete abolition of saddle slip andtransformation in work quality when lameness was eliminated bydiagnostic analgesia, emphasising that saddle slip may highlight thepresence of low-grade and subclinical lameness. This indicates that evensubtle lameness can influence a horse’s entire way of moving. One horseridden by an experienced rider, examined 2 months prior to thecommencement of the study, had no overt lameness, but had grade 2saddle slip and was presented for that reason. Although no overt lamenesscould be seen, the horse’s performance improved substantially followingintra-articular analgesia of the tarsometatarsal joint, which abolished thesaddle slip completely (S. J. Dyson, unpublished data).

Our findings demonstrated that saddle slip can also occur because of anasymmetrical (or ill-fitting) saddle. In 2 horses, the saddle slip wasunchanged after lameness was abolished by diagnostic analgesia, whichmakes it highly unlikely that it was induced by lameness. Both horses werealso ridden with correctly fitting saddles and no saddle slip was apparent.Although saddle slip is often a manifestation of hindlimb lameness, it canalso occur due to asymmetrical muscling of the horse’s back, asymmetricalposition of the rider [3] or asymmetry of the biomechanical function of thehorse, but based on our results these causes are less common thanhindlimb lameness.

Only 13 riders of 39 horses with saddle slip had recognised saddle slipprior to the clinical investigation. However, in all except one horse withsaddle slip, riders from the Animal Health Trust were able to feel saddle slip.Better education of riders may be required to recognise saddle slip,especially because it may be an important indicator of the presence oflameness. Five of the normal riders in the present study sat crookedly andcould have induced saddle slip; however, a similar degree of saddle slip wasseen when ridden by riders from the Animal Health Trust, who all satsquarely when riding a variety of different horses without saddle slip.

Some movement of the saddle can potentially be detected in horses freefrom lameness. In 7 Grand Prix dressage horses free from lameness riddenon a treadmill in rising trot, there was saddle rotation around the verticalaxis away from the supporting hindlimb [20]. Alternation of the forcebetween right and left was also seen in the trajectory of the COP [19]. In thepresent study, the saddle slip caused by hindlimb lameness was obvious

and occurred with at least 2 riders. The saddle slip was consistently morepronounced with a lighter weight rider, despite the lameness being lessobvious, presumably because a lighter rider is less able to stabilise andmaintain a straight position of the saddle. However, when saddle slip wasrelated to an asymmetrical saddle, slip was greater with a heavier weightrider. This observation was based on only 2 horses, but mirrors previousobservations (S. J. Dyson, unpublished data). A heavier rider mayexacerbate the uneven pressure caused by the unequal thickness andshape of the panels of the saddle. We also observed that the saddle slipassociated with lameness was much less in sitting trot or when ridingwithout stirrups compared with rising trot. This may be because in risingtrot the rider induces some rotation, which exacerbates sliding of thesaddle. Riders impose more force in the stirrups in rising trot comparedwith sitting trot [31], although there were no significant differences in peakforces under the saddle in rising and sitting trot [32]. However, when forceson the back were calculated using kinematics of a rider, there was asignificant difference in peak vertical force between rising and sitting trot[22]. In the present study, a heavier rider was more effective than a lighterrider in maintaining stability of the saddle in sitting trot. However, in cantersaddle slip was sometimes greater than in trot, probably related to theasymmetrical 3-beat gait, which impairs the rider’s ability to stabilise thesaddle despite sitting.

Saddle slip associated with hindlimb lameness was increased in circlescompared with straight lines. Low-grade hindlimb lameness is oftenaccentuated when a horse is ridden, especially in circles of approximately10 m diameter [8]. Assessment of movement of the trunk and tubera coxaein horses free from lameness was compared in straight lines and in circleson the lunge in unridden horses using inertial measurement units [33].There were significant repeatable patterns of asymmetry, which potentiallymay be magnified in the face of lameness and alter forces on the saddle andrider. Based on the results of the present study, there is clearly a complexinteraction between the size of a circle and its effect on the degree ofhindlimb lameness and its characteristics, and whether the lameness isworse on the outside or inside of a circle, and saddle slip.

We demonstrated that the shape of the horse’s back had an influence onsaddle slip, with ‘rounder’ horses at the level of T13 being more at risk. Thismay reflect greater dorsoventral displacement at T13, which appears to bethe site that comes closest to the lowest back line. The maximal range ofvertical displacement occurs at this site, where the equine body centre ofmass is aligned with the rider’s centre of mass [34,35]. However, there wasno association between the BCS and saddle slip. A large proportion of thehorses in the study were BCS 3 [25]. Use of a more sensitive method ofcondition scoring, such as the 9-point scale [36], may have permittedfurther differentiation among horses within BCS 3.

In our study, the saddle usually slid towards the side of the lamerhindlimb, but sometimes towards the less lame or nonlame limb. Whensaddle slip occurred towards the side of the lamer limb, this was usuallyassociated with the lamer limb on the outside of a 20 m circle. In contrast,when the saddle slid towards the sound or less lame hindlimb, this was withthe lamer limb on the inside of a circle. We need to understand better thebiomechanical changes in gait that occur to cause these 2 differentpatterns of saddle movement and to establish the interrelationshipbetween the symmetry and amplitude of equine in vivo back movement,the saddle movement and lameness in a variety of movement conditions.Combining conventional techniques for lameness assessment with backkinematics data using inertial measurement units mounted on the saddle,the rider and the horse [35] and force measurements under the saddle [37],alongside measurements of back muscle activity using electromyography[38,39] and ultrasonographic evaluation of multifidus [40], may help toimprove basic understanding of vertebral biomechanics and the changesthat occur with lameness.

The Flexible Curve Ruler is widely used in the saddle-fitting industry andhas been considered to be a reliable tool. However, there has been noscientific validation prior to the present study. Our data indicate that usingstrict consistency in technique (the horse standing squarely on ahorizontal, level surface, acquiring measurements before exercise atreliably identified predetermined sites), accurate, repeatable data can beobtained. The ruler is prone to fatigue, and regular replacement isrecommended to ensure consistency of results. However, the potentialconsequences of different patterns of work prior to back shape



measurements have yet to be investigated. We also need to determine thespeed with which body shape may change over time, influenced by typeand intensity of work, skeletal maturity, work discipline and nutrition.

We have previously demonstrated that hindlimb lameness can causeback stiffness [9]. Following diagnostic analgesia to abolish lameness, therewas increased back mobility, assessed both subjectively and objectivelyusing a Pliance pressure mat. The present study provided further evidencethat following diagnostic analgesia to eliminate hindlimb lameness, therewas an improved range of movement through the horse’s back andabolition of saddle slip. This was also confirmed objectively in 3 horses byusing a pressure mat, which demonstrated increased cranial–caudalexcursion of the COP and decreased sideways movement of the COP, andthe difference in pressure between the lame and sound diagonal stancephases disappeared.

Limitations of the studyThis study had some limitations. The horses were a referral populationand therefore may not be typical of the general horse population, in whichthere may be greater occurrence of musculoskeletal asymmetries. Theprevalence of crooked riders and saddle-fitting problems may notnecessarily reflect the general population. The study was not performedblinded, which could potentially have introduced bias. There were only 39horses with saddle slip in the study, in 37 of which saddle slip was caused byhindlimb lameness. This was not enough to allow investigation of therelationships between the grade and direction of saddle slip and the sourceof pain causing lameness. We need to gain a better understanding ofthe biomechanical changes that occur with lameness that inducesaddle slip. Combining clinical diagnosis with inertial measurement unitmeasurements in a larger number of horses would enable us to establishpatterns related to the site of pain causing lameness, back movement andsaddle slip. Although there was no relationship between the presence ofsaddle slip and the degree of lameness in the present study, horses thatwere severely lame in hand were not ridden and were therefore notincluded in the study. This may potentially have influenced the results. Alarge proportion of horses had bilateral hindlimb lameness; it is difficultto grade bilateral lameness accurately [27]. One horse ridden by aprofessional rider had grade 2 saddle slip towards the lame hindlimbabolished by diagnostic analgesia. However, the horse was assessed byonly one rider due to its potentially dangerous behaviour and was notincluded in the study.

The back shape measurements were obtained in standing horses frompredetermined anatomical locations and therefore do not represent theequine back shape in its entirety and in a dynamic situation. By combiningthe Flexible Curve Ruler measurements with 3-dimensional back shapescanning Artecd, alongside Horse Weight Measuring Tapee, it may bepossible to evaluate the back in its entirety and also to measurequantitatively the change in back shape that may occur with work and overtime. This was a single-point study; however, most horses wereinvestigated over 2–3 days, and the presence or absence of saddle slipwere consistent features. A proportion of horses were re-evaluated duringthe study period after treatment of lameness. In those horses in whichlameness resolved, saddle slip invariably disappeared also. There wereonly 2 horses in which saddle slip was associated with an ill-fitting saddle,but the findings recorded are consistent with previous observations(S. J. Dyson, unpublished data).

ConclusionsBased on the present study, there is evidence that saddle slip occurs in ahigh percentage of horses with hindlimb lameness. Moreover, saddle slipmay highlight the presence of low-grade and subclinical hindlimblameness. Saddle slip is usually blamed on saddle fit and horse shape. Ourfindings emphasise the need for education of owners, veterinarians,physiotherapists, trainers, riders and saddle fitters that saddle slip isfrequently an indicator of lameness, not necessarily a manifestation of anill-fitting saddle or asymmetric shape of the horse’s back. It appears thatsaddle slip as a manifestation of hindlimb lameness has previously beenunderestimated. In many horses hindlimb lameness goes unrecognised,and early recognition of lameness is important for appropriate treatmentand rapid return to work. Detection of saddle slip provides an opportunity

for the owner, riders and trainers to detect low-grade and subclinicallameness, with important welfare consequences.

Authors’ declaration of interests

There are no conflicts of interest.

Source of funding

There are no sources of funding.

Acknowledgements

We thank the many referring veterinary surgeons without whom this studywould not have been possible, the Saddle Research Trust and BiosenseMedical.

Authorship

Both authors contributed to all parts of the study.

Manufacturers’ addressesaAti 90 cm, Taiwan.bTekscan, Boston, Massachusetts, USA.cSPSS Inc., Chicago, Illinois, USA.dSteinbichler Scanning Technology, Bromsgrove, UK.eWINtape, China.

References

1. Harman, J. (1999) Tack and saddle fit. Vet. Clin. N. Am.: Equine Pract. 15,247-261.

2. Lagarde, J., Kelso, J.A., Peham, C. and Licka, T. (2005) Coordination dynamics ofthe horse-rider system. J. Mot. Behav. 37, 418-424.

3. Symes, D. and Ellis, R. (2009) A preliminary study into rider asymmetry withinequitation. Vet. J. 181, 34-37.

4. Clayton, H.M., Kaiser, L.J., dePue, B. and Kaiser, L. (2011) Centre of pressuremovements during equine assisted activities. Am. J. Occup. Ther. 65, 211-216.

5. de Cocq, P., van Weeren, P.R. and Back, W. (2006) Saddle pressure measuring:validity, reliability and power to discriminate between different saddle-fits. Vet.J. 172, 265-273.

6. Meschan, E.M., Peham, C., Schobesberger, H. and Licka, T.F. (2007) Theinfluence of the width of the saddle tree on the forces and the pressuredistribution under the saddle. Vet. J. 173, 578-584.

7. Dyson, S. (2010) Poor performance and lameness. In: Diagnosis andManagement of Lameness in the Horse, 2nd edn., Eds: M.W. Ross and S.J.Dyson, Elsevier, St Louis, Missouri. pp 920-925.

8. Dyson, S. (2012) How to recognise hindlimb lameness: an obvious lameness tosubtle gait abnormalities. In: Proceedings of the American Association EquinePractitioners Focus Meeting on Hindlimb Lameness, Oklahoma City. CD Rom,pp 2-9.

9. Byström, A., Dyson, S. and Murray, R. (2011) Saddle pressure patterns arealtered by diagnostic analgesia in ridden horses with poor performance.Proceedings 50th British Equine Veterinary Association Congress, EquineVeterinary Journal Ltd, Fordham. p 181.

10. Audigié, F., Pourcelot, P., Degueurce, C., Denoix, J.M. and Geiger, D. (1999)Kinematics of the equine back: flexion-extension movements in sound trottinghorses. Equine Vet. J. 30, 210-213.

11. Johnston, C., Holm, K.R., Erichsen, C., Eksell, P. and Drevemo, S. (2004)Kinematic evaluation of the back in fully functioning riding horses. Equine Vet. J.36, 495-498.

12. Buchner, H.H., Savelberg, H.H., Schamhardt, H.C. and Barneveld, A. (1996)Head and trunk movement adaptations in horses with experimentally inducedfore- or hindlimb lameness. Equine Vet. J. 28, 71-76.

13. Pourcelot, P., Audigié, F., Degueurce, C., Denoix, J.M. and Geiger, D. (1998)Kinematics of the equine back: a method to study the thoracolumbarflexion-extension movements at the trot. Vet. Res. 29, 519-525.



14. Gómez Alvarez, C.B., Wennerstrand, J., Bobbert, M.F., Lamers, L., Johnston, C.,Back, W. and van Weeren, P.R. (2007) The effect of induced forelimb lamenesson thoracolumbar kinematics during treadmill locomotion. Equine Vet. J. 39,197-201.

15. Gómez Alvarez, C.B., Wennerstrand, J., Bobbert, M.F., Lamers, L., Johnston, C.,Back, W. and van Weeren, P.R. (2008) The effect of induced hindlimb lamenesson thoracolumbar kinematics during treadmill locomotion. Equine Vet. J. 40,147-152.

16. Wennerstrand, J., Johnston, C., Roethlisberger-Holm, K., Erichsen, C., Eksell, P.and Drevemo, S. (2004) Kinematic evaluation of the back in the sport horse withback pain. Equine Vet. J. 36, 707-711.

17. Wennerstrand, J., Gómez Alvarez, C.B., Meulenbelt, R., Johnston, C., vanWeeren, P.R., Roethlisberger-Holm, K. and Drevemo, S. (2009) Spinalkinematics in horses with induced back pain. Vet. Comp. Orthop. Traumatol.22, 448-454.

18. Zimmerman, M., Dyson, S. and Murray, R. (2011) Close, impinging andoverriding spinous processes in the thoracolumbar spine: the relationshipbetween radiological and scintigraphic findings and clinical signs. Equine Vet. J.44, 178-184.

19. Fruehwirth, B., Peham, C. and Schobesberger, H. (2004) Evaluation of pressuredistribution under an English saddle in walk, trot and canter. Equine Vet. J. 36,754-757.

20. Byström, A., Rhodin, M., von Peinen, K., Weishaupt, M.A. and Roepstorff, L.(2009) Basic kinematics of the saddle and rider in high-level dressage horsestrotting on a treadmill. Equine Vet. J. 41, 280-284.

21. de Cocq, P., Prinsen, H., Springer, N.C., van Weeren, P.R., Schreuder, M., Muller,M. and van Leeuwen, J.L. (2009) The effect of rising and sitting trot on backmovements and head-neck position of the horse. Equine Vet. J. 41, 423-427.

22. de Cocq, P., Duncker, A., Clayton, H., Bobbert, M., Muller, M. and van Leeuwen,J. (2010) Vertical forces on the horse’s back in sitting and rising trot. J. Biomech.43, 627-631.

23. Byström, A., Rhodin, M., von Peinen, K., Weishaupt, M.A. and Roepstorff, L.(2010) Kinematics of saddle and rider in high-level dressage horses performingcollected walk on a treadmill. Equine Vet. J. 42, 340-345.

24. Byström, A., Stalfelt, A., Egenvall, A., Von Peinen, K., Morgan, K. andRoepstorff, L. (2010) Influence of girth strap placement and panel flockingmaterial on the saddle pressure pattern during riding of horses. Equine Vet. J.42, Suppl. 38, 502-509.

25. Carroll, C.L. and Huntingdon, P.J. (1988) Body condition scoring and weightestimation of horses. Equine Vet. J. 20, Suppl. 1, 41-45.

26. Hopkins, W.G. (2000) Measures of reliability in sports medicine and science.Sports Med. 30, 1-15.

27. Dyson, S. (2011) Can lameness be reliably graded? Equine Vet. J. 43, 379-382.

28. Ferguson-Pell, M., Nicholson, G., Lennon, P. and Bain, D. (2001) Comparative ofpressure mapping systems (2): cushion testing. Proceedings of the 24thRehabilitation Engineering and Assistive Technology Society of North AmericaConference, Arlington, Virginia. pp 289-291.

29. von Peinen, K., Wiestner, T., von Rechenberg, B. and Weishaupt, M.A.(2010) Relationship between saddle pressure measurements and clinicalsigns of saddle soreness at the withers. Equine Vet. J. 42, Suppl. 38, 650-653.

30. Dyson, S. and Murray, R. (2003) Pain associated with the sacroiliac joint region:a clinical study of 74 horses. Equine Vet. J. 35, 240-245.

31. van Beek, F., de Cocq, P., Timmermann, M. and Muller, M. (2012) Stirrup forcesduring horse riding: a comparison between sitting and rising trot. Vet. J. 193,193-198.

32. Peham, C., Kotschwar, A., Borkenhagen, B., Kuhne, S., Molsner, J. and Baltacis,A. (2010) A comparison of forces acting on the horse’s back and the stability ofthe rider’s seat in different positions at trot. Vet. J. 184, 56-59.

33. Starke, S., Willems, E., May, S. and Pfau, T. (2012) Vertical head and trunkmovement adaptations of sound horses trotting in a circle on a hard surface.Vet. J. 193, 73-80.

34. Buchner, H.H., Obermuller, S. and Scheidl, M. (2001) Body centre of massmovement in the lame horse. Equine Vet. J. Suppl. 33, 122-127.

35. Warner, S.M., Koch, T.O. and Pfau, T. (2010) Inertial sensors for assessment ofback movement in horses during locomotion over ground. Equine Vet. J. 42,Suppl. 38, 417-424.

36. Henneke, D.R., Potter, G.D., Kreider, J.L. and Yeates, B.F. (1983) Relationshipbetween condition score, physical measurements and body fat percentages inmares. Equine Vet. J. 15, 371-372.

37. von Peinen, K., Wiestner, T., Bogisch, S., Roepstorff, L., van Weeren, P.R. andWeishaupt, M.A. (2009) Relationship between the forces acting on the horse’sback and the movements of rider and horse while walking on a treadmill.Equine Vet. J. 41, 285-291.

38. Groesel, M., Zsoldos, R.R., Kotschwar, A., Gfoehler, M. and Peham, C. (2010) Apreliminary model study of the equine back including activity of longissimusdorsi muscle. Equine Vet. J. 42, Suppl. 38, 401-406.

39. Licka, T., Peham, C. and Frey, A. (2004) Electromyographic activity of thelongissimus dorsi muscles in horses during trotting on a treadmill. Am. J. Vet.Res. 65, 155-158.

40. Stubbs, N.C., Hodges, P.W., Jeffcott, L.B., Cowin, G., Hodgson, D.R. andMcGowan, C.M. (2006) Functional anatomy of the caudal thoracolumbarand lumbosacral spine in the horse. Equine Vet. J. 38, Suppl. 36, 393-399.



an investigation of the relationship between hindlimb ... & dyson saddle slip... · an...

Documents