Evaluation of on-farm methods assessing locomotor disorders of individual or group housed sows

The aim was to evaluate methods for on-farm assessment of locomotor disorders in gestating sows, and the influence of housing system on the prevalence of locomotor disorders. Of the 311 sows studied across 10 farms, 120 sows were housed in groups of 4–8 sows per pen and 191 sows were housed in individual stalls. Data recorded were: 3-point gait score, walking speed (≤0.8, ≤1.25, >1.25 m/s), stride length (<83, <100, ≥100 cm), number of steps during 30s at feeding (<3, <6, ≥6), latency to lie down after feeding (≤30, ≤60, >60 min), ability to stand up after being stimulated (“stand-up test”: moving freely, hesitancy, refusal), number of feet with uneven toe size and bursitis (0, 1, >1), and presence of cracks (>1 cm) on the side wall. The percentage of sows in each category was calculated. Cochran-Mantel-Haenszel and Chi-squared tests were used to determine the association between variables and housing systems. Following gait scoring, 49% of sows were sound (score 0), 35.5% walked stiffly (score 1) and 15.5% were visibly lame (score 2). There were no differences between housing systems in the percentage of sows within each gait score (P > 0.05). The percentage of sows having feet with uneven toe size, bursitis or cracks on the side wall was higher in stalls than in pens (P < 0.05). No associations were found between the 3-point gait score and the number of steps during 30s, the stand-up test score or the latency to lie down after feeding (P > 0.05). The percentage of sows having reduced walking speed and stride length was higher for sows scored 2 than for sows scored 0, in stalls only (P < 0.05). Results indicated no differences in the prevalence of locomotor disorders between housing systems based on the gait score, but highlighted more foot problems in stalls than in pens. New quantitative indicators such as stride length and walking speed seem to be good indicators of locomotor disorders for sows housed in stalls but not in pens. This suggests that methods used to detect lameness are not always applicable from one housing system to another. Other factors associated with the housing system such as the possibility to exercise may interfere with the assessment.