Birth weight threshold for identifying piglets at risk for preweaning mortality

Abstract Several studies have suggested there is a critical relationship between piglet birth weight and preweaning mortality. Thus, the objective of the current work was to identify a birth weight threshold value for preweaning mortality. Birth weight and survival data from two studies involving a combined total of 4,068 piglets from 394 litters on four commercial farms (three European, one U.S.) were compiled for a pooled, multistudy analysis. Overall preweaning mortality across the two studies was 12.2%. Key variables used in the analysis were piglet birth weight (measured within 24 h of birth) and corresponding survival outcome (dead or live) by weaning at 3–4 wk of age. A mixed effects logistic regression model was fit to estimate the relationship between preweaning mortality and birth weight. A random effect of study was included to account for overall differences in mortality between the two studies. A piecewise linear predictor was selected to best represent the drastic decrease in preweaning mortality found as birth weight increased in the range of 0.5–1.0 kg and the less extreme change in weight above 1.0 kg. The change point of the birth weight and preweaning mortality model was determined by comparing model fit based on maximizing the likelihood over the interval ranging from 0.5 to 2.3 kg birth weight. Results from the analysis showed a curvilinear relationship between birth weight and preweaning mortality where the birth weight change point value or threshold value was 1.11 kg. In the combined data set, 15.2% of pigs had birth weights ≤1.11 kg. This subpopulation of pigs had a 34.4% preweaning mortality rate and represented 43% of total preweaning mortalities. These findings imply interventions targeted at reducing the incidence of piglets with birth weights ≤1.11 kg have potential to improve piglet survivability. Additional research is needed to validate 1.11 kg as the birth weight threshold for increased risk of preweaning mortality.