Our objective was to quantify the associations between climatic factors and a death occurrence of peripartum pigs from 16 to 19 weeks after successful service during hot and humid or cold seasons. The study used lifetime records of 93,837 females entered into 98 Japanese commercial herds from 2003 to 2007. The climate data were obtained from 21 weather stations close to the studied herds. Average daily maximum (HT) and minimum temperature (LT), and relative humidity for week 15 of gestation for each pregnant pig were coordinated with the respective pig's performance data. Multilevel logistic regression models were applied to two of the three separate datasets. One dataset included females due to farrow during the hot and humid season (June-September), and another comprised females due to farrow during the cold season (December-March). Of the 8381 females that died throughout the year, 11.5% of pregnant pigs died between 16 and 17 weeks after service, and 44.3% of farrowed females subsequently died from 16 to 19 weeks after service. Mean (ranges) HT in the hot and humid season and LT in the cold season were 28.7(13.4-39.8)degrees C and 1.6 (-14.8 to 17.6)degrees C, respectively. Means of relative humidity in the hot and humid season and cold season were 73.6 (35-98)% and 64.9 (21-99)%, respectively. In the hot and humid season, a higher HT was associated with a higher occurrence of death for parity 0-1 females (P < 0.05), but not for parity 2 or higher sows (P >= 0.38). The odds ratio was 1.030 (95% confidence intervals: 1.005-1.056) for HT in parity 0-1 females. Also, higher relative humidity was associated with a higher occurrence of death for parity 0-3 females (P < 0.05), but not parity 4 or higher sows (P >= 0.21). In the cold season, a higher occurrence of death of parity 4 or higher sows was associated with lower LT (P < 0.05). Also, the occurrence of death of parity 6 or higher sows was associated with higher relative humidity in the cold season (P < 0.05). For parity 0-3 females, there were no associations between the occurrences of death and either LT or relative humidity during the cold season (P >= 0.11). Therefore, it is recommended to install cooling systems and thick insulation to prevent increases in occurrences of pig deaths due to HT or LT. (C) 2014 Elsevier B.V. All rights reserved.
