Animals and housing
An observational prospective study was carried out in a Danish pig herd. Sixty randomly selected crossbred Landrace-Yorkshire (LY) sows from the herd were included. The sows were tethered during the gestation, with concrete floor in the lying area, and slatted floor in the dunging area. The farmer decided exclusively when to cull the sows, which did not differ from usual procedures. The time of culling was recorded and varied from first to ninth parity.
Gait and posture scoring
The gaits and postures variables, which were often bilateral, were scored before first mating and after every farrowing until culling. The variables were defined according to earlier publications [9,10]. The scoring procedure was performed by one observer outside the pen with the animal in motion. The following 11 variables of the gait and posture, of which buck-kneed forelegs, fore and hind legs turned out, and stiff in front and rear have been shown to be associated with osteochondrosis and arthrosis , were scored on a scale from 1 (normal) to 5 (severe):
• Buck-kneed forelegs
• Forelegs turned outwards
• Upright pasterns forelegs
• Weak pasterns forelegs
• Standing under position hind legs
• Hind legs turned outwards
• Steep hock joint
• Weak pasterns hind legs
• Stiff in front
• Stiff in rear
• Swaying hindquarters
Elbow and knee joints were collected at slaughter. Complete sets of joints were obtained from 33 animals, while incomplete sets were sampled from 27 sows. In these cases the following materials were missing: left radius and ulna (one sow); right elbow joint (9 sows); left (20 sows) and right (24 sows) knee joint.
All joints were opened and evaluated macroscopically in specified locations: (I) the medial humeral condyle, (II) the lateral humeral condyle, (III) fovea capitis radii, (IV) incisura trochlearis of ulna, (V) processus anconeus of ulna, (VI) the medial femoral condyle, and (VII) the lateral femoral condyle. The locations were assesed for the presence of: (a) erosions, (b) ulcerations, (c) repair reactions, (d) marginal osteophytes, and (e) infolding of the joint cartilage according to a template (Fig. 1a–f) and scored as normal (0), moderate (1), when the lesion involved less than 20% of the articular surface or severe (2), when the lesion exceeded 20% of the articular surface.
Figure 1 Template for categorizing macroscopical joint lesions in sows. a: Cartilage erosion (arrows) on the medial humeral condyle. b: Cartilage ulceration (arrow) on the medial femoral condyle. c: Cartilage repair (arrow) of the medial femoral condyle d: Marginal osteophytes (arrows) on processus anconeus of ulna. e: Cartilage infoldings (arrow) on the medial femoral condyle. f: Cartilage infoldings on the medial femoral condyle. Cross section of Fig. 2e.
In order to confirm the nature of the macroscopical lesions, a representative number of the specified joint lesions was evaluated histologically according to a template (Fig. 2a–d) and according to the following definitions: (I) erosion: thinning and loss of the surface cartilage, (II) ulceration: the articular cartilage was lost and the subchondral bone was exposed, including flap formation in osteochondritis dissecans lesions, (III) repair: defect in the cartilage substituted by fibrous tissue or fibrocartilage, (IV) osteophytes: formation outside the bone consisting of osseous trabeculae, and (V) infolding: articular cartilage was protruding into the subchondral bone.
Figure 2 Template for histological classification of joint lesions in sows. a: Superficial cartilage erosions (arrows) of variable thickness are present. Articular cartilage of the medial humeral condyle. Haematoxylin and eosin. Bar = 100 μm. b: Typical osteocondrotic lesion in the form of osteochondritis dissecans (arrow). Articular cartilage of the lateral humeral condyle. Haematoxylin and eosin. Bar = 200 μm. c: Fibrous tissue and fibrocartilage are filling out a defect of the articular cartilage. Articular cartilage of the medial humeral condyle. Haematoxylin and eosin. Bar = 125 μm. d: Infoldings of thickened (retained) articular cartilage are present (arrows). Articular cartilage and subchondral bone of the medial humeral condyle. Masson's Trichrome. Bar = 5 mm.
The PROC CORR procedure in SAS was used for analysis, and the mutual correlations between similar lesions of left and right side were analysed. The same procedure was used for analysing the mutual correlation between lesions in the same joint, one side at a time.
The frequencies of scorings of the gait and posture variables were analysed. The associations between the gait and posture variables and the joint lesions were analysed one at a time by using the maximum score over time of the 11 variables for each sow against all the joint lesion scores. A backward elimination procedure was used by removing one variable at a time with highest P-value until only variables with a P-value below 0.5 were left in the model. The frequencies of the joint lesions were examined and lesions observed in less than 10% of the animals were eliminated from the analyses. The procedure PROC GLM in SAS was used to estimate Pearson correlation coefficients.