Defining osteochondrosis and its genetic background · 2010-11-26 · Defining osteochondrosis and its genetic background Ilse van Grevenhof Animal Breeding and Genomics Centre Bart

Defining osteochondrosis and its

genetic background

Ilse van GrevenhofAnimal Breeding and Genomics Centre

Bart Ducro

Piter Bijma

René van Weeren

Hans van Tartwijk

KWPN

What is OC?

Health problem in horses

Disturbance of ossification

In joints of young growing animals

Previous studies show prevalences vary 7% – 64%

Definition OC

Number of joints, radiographs, locations

Preselected data sample

Aims

Assess prevalences Joint level

Animal level

Stallion level

Study relationship Left and right joint

Flattenings and fragments

Genetic parameters Heritabilities

Genetic correlations

Materials

Dutch Warmblood population

2005/ 2006

OC was scored in 811 yearlings

After ‘point of no return’ (~12 months)

Randomly selected sample

Yearlings descended from 32 stallions

22 - 28 yearlings per stallion

Young and old stallions

Dressage and show jumping stallions

Methods

Scored on categorical scale A to E (Dik et al., 1999)

A = Ideal

B = Smooth flattened bone structure

C = Irregular flattened bone structure

D = Small bony fragment

E = Large bony fragment

Material and methods

OC scored on: 8 joints

14 radiographs

28 locations

Stifle Left 5 locations

Right 5 locations

Hock Left 7 locations

Right 7 locations

Fetlock 1 location

Methods: transformation to continuous scale

Aims: Estimate heritabilities and relationships

Problem: Observations are categorical

Needed: Transformation of categorical to continuous scale

Method: Based on prevalences

Result: 3 continuous traits FLAT

FRAG

ALL (FLAT + FRAG)

For each joint

For the entire animal (all joints combined)

Results

Animal level

Worst animal had 11 out of 28 ‘ideal’ locations

Best animal had 28 out of 28 ‘ideal’ locations

Joint Percentage of animals with fully ‘ideal’ joints

Stifle 61%

Hock 69%

Fetlock 65%

Total 30%

Results

Sire level

Mean: 30% of yearlings show fully ‘ideal’ joints

Worst stallion had 12% of yearlings with fully ‘ideal’ joints

Best stallion had 58% of yearlings with fully ‘ideal’ joints

No difference between show jumping/ dressage stallions

Are there prospects for breeding?

Sires (n=32)

Phe

noty

pic

SD

Standardized EBV of sires (All joints combined)

-0.470.33

~58%

~12%

Cluster analysis

Relationships

Reduction of variables

Based on correlation matrix

ALLHock

Front

Hind

Fetlock

Flat

Flat

Stifle

Frag

Front

Hind

Flat

FragR

L

R

L

R

Frag

L

R

L

R

L

L

L

R

L

R

R

1.0 0.20.6Proportion of variance explained

Cluster analysis

Clustering order:

Left and right joints most similar

FLAT and FRAG less similar

Joints least similar

Reduction from 16 to 6 variables

~0.63

Conclusions

Prevalences (% not ideal)

Varies between joints

Cluster analyses indicated:

Left and right joints can be combined in analyses

FLAT and FRAG should be separated in analyses

Large differences between stallions

Results hard to compare with literature

Definition OC

Number of joints, radiographs, locations

Statistical model

OC = µ + sex + vet + age + year + animal + e

OC is the continuous FLAT, FRAG or ALL for a joint or the entire animal

REsidual Maximum Likelihood (REML)

Pedigree N = 7799

Heritabilities

Joint h2 ALL (SE)

h2 FLAT (SE)

h2 FRAG (SE)

Animal 0.23 (0.09) 0.08 (0.06) 0.22 (0.09)

Stifle joint 0.05 (0.05) 0.07 (0.06) 0.02 (0.04)

Hock joint 0.36 (0.11) 0.15 (0.08) 0.26 (0.09)

Fetlock

joints

0.14 (0.08) 0.08 (0.10) 0.06 (0.07)

Heritabilities

Do we need to score left and right joints?

Heritabilities of left and right joints were highly similar

Combined h2 is higher (increase with 16%)

Do we need to score from A-E or will binary (0/1) do?

h2 entire animal: 0.23 0.15

Genetic correlations

Genetic correlation in entire animal

FLAT and FRAG was 0.80 (± 0.25)

Genetic correlations between joints

Stifle and hock joints: 0.59

Fetlock and other joints: 0.09 - 0.26

Which joints are relevant for breeding?

Stifle joint

High prevalence

High clinical relevance (low prognosis surgical treatment)

High SE for genetic correlation with hock joint (0.59 0.37)

Hock joint

High prevalence of fragments

High heritable

Fetlock joint

Intermediate heritable

Low genetic correlations (no correlated response)

Are both FLAT and FRAG relevant for breeding?

FLAT and FRAG

Although rg between FLAT and FRAG is high (0.80)

SE is large (0.25)

Heritabilities are substantially different

What can be concluded?

For breeding, OC should be recorded in:

All 4 joints

Flattenings and fragments

Left and right joints

Scored in >2 categories

Linear OC value

0

0,1

0,2

0,3

0,4

-3 -2 -1 0 1 2 3

y-axis: probability

x-axis: standard deviation

part belonging to

percentage of

category A

A

B

CD E

-0.23

Locations Stifle

1. Lateral femoral trochlea

2. Medial femoral trochlea

3. Sulcus distal femur

4. Patella

5. Other locations

Hock 1. Sagittal ridge of distal tibia

2. Lateral trochlea of talus

3. Medial trochlea of talus

4. Lateral malleolus of tibia

5. Medial malleolus of tibia

6. Base of talus

7. Other locations

Fetlock Proximodorsal part of the sagittal ridge of 3rd metacarpal/metatarsal bone

Radiographs

Stifle

Lateromedial

Anterior-posterior lateral oblique

Hock

Anterior-posterior

Lateromedial

Anterior-posterior lateral oblique

Fetlock

Lateromedial