Joint lesions are a major cause of euthanasia and culling of sows in Denmark and are of importance both economically and in relation to animal welfare 1 . Joint lesions of sows are frequent causes of leg weakness, and non-inflammatory joint diseases as arthrosis and osteochondrosis are main causes of lameness 2 3 4 . Osteochondrosis developes in growing animal and is due to a failure in the endochondrale ossification of the articular cartilage and the growth plate 5 . The lesions caused by osteochondrosis can heal completely 2 or progress into secondary arthrosis in the adolescent animal 5 . The aetiology of osteochondrosis is thought to be multifactorial, and trauma, heredity, rapid growth, nutrition, and anatomical conformation are factors associated with this disease 5 6 7 . Non-osteochondrosis-related arthrosis (i.e. primary arthrosis) is characterized by fibrillation and ulceration of the articular cartilage and of eburnation of the subchondral bone 5 . The pathogenesis of primary arthrosis of sows is not well understood, but the confinement of sows and the subsequent limitations of exercise have been suggested as a possible aetiology 8 . Osteochondrosis and arthrosis in sows are often bilateral and symmetrical and are frequently observed in the distal humerus and femur 2 .

Focus on the association between clinical observations and lesions of the locomotive system has been the objective in only a few porcine studies 3 4 . Therefore, it is uncertain which specific joint lesions actually are associated with the different types of abnormal gait and posture in pigs. The clinical examination of sows has until now been of limited use when trying to asses the cause of lameness, and a post-mortem examination of the animal has been preferred to differentiate the various causes of lameness 3 .

The present study was performed in order to examine the correlations between certain joint lesions and defined gait and posture variables in sows.

Joint lesions were observed more often in the elbow joint compared to the knee joint (Tables 1 and 2). The most frequent lesion in the elbow joint was erosion of the articular cartilage, in particular on the medial humeral condyle (left side 95%, and right side 84%). Also ulceration (left side 18%, right side 10%) and repair (left side 13%, right side 16%) of the articular cartilage of the medial humeral condyle, as well as formation of marginal osteophytes of processus anconeus (left side 14%, right side 16%) of ulna were often observed. In the knee joint, erosion (left side 15%, right side 42%) and ulceration (left side 10%, right side 6%) of the articular cartilage of the medial femoral condyle were noted as the most frequent lesions.

Because a significant correlation between similar lesions of the left and the right side (from r = 0.25 to r = 0.71) was found, the two sides were subsequently pooled.

The mutual correlations between lesions within the joints (Tables 3 and 4) showed a strong correlation between erosions in the lateral condyle of humerus and cartilage erosion of incisura trochlearis on ulna (P < 0.001) and between erosions in the lateral condyle of humerus and marginal osteophytes on the processus anconeus of ulna (P < 0.001). However, no correlations were seen between the same types of lesions in the medial condyle of humerus. Also a strong correlation between cartilage erosion of fovea capitis on radius and cartilage erosion of incisura trochlearis on ulna was observed (P < 0.001). In the knee joints a strong correlation between erosion and ulceration in the medial condyle of femur was registered (P < 0.01).

The scorings of the variable 'stiff in rear' and 'swaying hindquarters' showed that 44% and 39% of scorings, respectively, were between 3 and 5 (Fig. 3).

The highest degree of positive associations were between 'hind legs turned out' and repair of the articular cartilage of the medial femoral condyle (P < 0.001) and with marginal osteophytes of the fovea capitis on radius (P < 0.01), and 'weak pasterns forelegs' and with marginal osteophytes of the fovea capitis on radius (P < 0.001) (Tables 5 and 6). 'Forelegs turned out' were positively associated with erosions of incisura trochlearis on ulna (P < 0.05). Moreover, significantly positive associations between 'stiff in front and in rear' and ulceration of the cartilage of the lateral humeral condyle were observed (P < 0.05). Significantly negative associations were found between 'weak pasterns on forelegs' and cartilage ulceration of the medial humeral condyle (P < 0.05) and cartilage infoldings of the medial femoral condyle that were verified to be of osteochondrotic origin (P < 0.01). A negative association was also found between 'stiff in rear' and cartilage erosion of radius (P < 0.05) and cartilage ulceration of the medial femoral condyle (P < 0.01).

Associations to the first and the last scoring were examined, too, but did not influence the results. No significant effect of parity was found.

Correlations between various lesions on the same articular surfaces and between lesions of opposing articular surfaces in the elbow and knee joints were observed. It was not obvious from the correlations which types of lesions preceded the other ones. However, because histology revealed erosions of the articular cartilage without ulcerations (Fig. 2a), it was most likely that erosions preceded ulcerations. An exception from this was in cases of osteochondritis dissecans, where ulceration was seen without erosion being present (Fig. 2b).

In accordance with results obtained in a previous study 4 , a correlation between erosion in the articular cartilage of the lateral humeral condyle and the presence of marginal osteophytes on processus anconeus of ulna was observed. The presence of marginal osteophytes was always observed together with erosion of the articular cartilage. By contrast, erosions were often seen without marginal osteophytes. Therefore, it is likely that cartilage lesions precede the formation of marginal osteophytes. However, in humans osteophytes may be present without any affection of the cartilage 12 , and it is assumed to be an adaptive and stabilizing reaction caused by instability of joints 13 . Therefore, it could be speculated that both cartilage lesions and osteophytes in sows are caused by joint instability.

The positive association between forelegs that are turned out and stiff movements of the front and rear legs and cartilage lesions in the elbow joint is in agreement with the results obtained by Jørgensen 9 . Weak pasterns on forelegs were both negatively and positively associated with lesions in the elbow and knee joints, but in particular marginal osteophytes on radius were positively associated (P < 0.001). The presence of weak pasterns on forelegs has previously been found to be positively associated with normal, brisk gait and negatively associated with osteochondrosis/osteoarthrosis 9 . However, in contrast to a study by Jørgensen (9), in which only lesions in the knee joint had an impact on hind legs being turned out, it was found that also lesions in the elbow joint were associated with this abnormal posture.

In the present study it was shown that some defined gait and posture variables reflected specific joint lesions in sows. Presence of 'stiff in front and rear legs' and 'forelegs turned out' were highly indicative of osteochondrotic and arthrotic lesions in the elbow joint. These observations could be helpful in the selection procedure of breeding animals and should encourage farmers to include animals with a low incidence of osteochondrosis in breeding programmes. However, further studies are needed to further strengthen and optimize the diagnostic tool.

Moreover, it was found that correlations between certain articular lesions exist.

The author(s) declare that they have no competing interests.

BJ designed the study and developed the gait and posture scoring methods. RKK examined and scored the joint lesions with assistance from HEJ. RKK and BJ performed the statistical analysis and RKK, BJ, and HEJ drafted the manuscript. All authors read and approved the final manuscript.