Factors associated with specific health, welfare and reproductive performance indicators in pig herds from five EU countries.

Production diseases are often of multi-factorial origin in which environment (housing, nutrition and management) health and reproductive challenges show complex interactions. The aim of this study was to identify specific environment-related factors and to discuss their associations with health, welfare and reproductive performance in sows and piglets, in diverse systems using data from 130 farms from five EU countries. Two sets of data were used: a) a questionnaire was developed for sows and piglets covering farm management, interventions and housing and b) farm production data covering various performance parameters. Eight parameters were further selected, four of which were related to sow reproductive performance (litter index, replacement rate, repeat breeding (i.e. failure to breed after one mating), weaning to first mating interval) and the remaining four to litter / piglet health performance (piglets born alive per litter, piglets born dead per litter, preweaning mortality rate and weaned piglets per litter). Univariable and multivariable linear models were employed to identify risk factors. Associations were considered significant if P ≤ 0.007 (a criterion of p ≤ 0.05 corrected for the number of parameters tested). Various risk and protective factors were identified for each tested outcome variable. Country effects were included in all models as a fixed factor. Adjusted R-squared values for the multivariable models varied between 9.6% (preweaning mortality) and 66% (litter index). Litter index (litters/sow/year) was negatively associated with a higher weaning age of the piglets. Housing recently weaned sows to be inseminated in a separate unit from the gestation unit had a positive association with litter index. Repeat breeding was negatively associated with PRRS-free farms, farms that bred (raised) all gilts on the farm and farms that perform farrowing induction of sows. PRRS-free farms were also associated with a higher replacement rate. Farms that bred gilts on the farm and PRRS-free farms were negatively associated with preweaning mortality. Natural ventilation in the gestation unit was associated with fewer piglets born alive and with fewer weaned piglets. Closed type of farms was associated with less piglets born dead. The use of open box housing system for pregnant sows (provision of individual resting areas) was associated with more weaned piglets. In conclusion, several factors related to applying good farm and health management, and optimal housing conditions showed positive association with various sow and piglet performance parameters. Further studies will help to assess causal links for these factors.