Predicted essential fatty acid intakes for a group of dairy cows also apply at individual animal level

•Dairy cows show inter-individual variations in essential fatty acid metabolism.•Group predictions of essential fatty acid intake were tested at individual level.•Dietary and animal factors explained the gap between predicted and observed values.•Correction of fatty acid intake using animal-only factors gave good results.•Group equations adjusted to dietary and animal factors can serve to predict intake. The ruminant requirements for essential fatty acids (EFA), particularly linoleic acid (LA) and alpha-linolenic acid (ALA), have not been fully determined, although evidence suggests that an adequate supply of polyunsaturated fatty acids (FA) could improve immunity and reproduction in transition cows. In previous studies we predicted EFA intake for a group of cows based on animal characteristics and milk EFA secretions. However, to support precision livestock feeding, we need to match the nutrient requirements and intakes of each cow as closely as possible. Our group-level predictions may not be accurate enough to estimate the EFA intake of an individual cow, due to inter-individual variations in EFA digestion and metabolism related to differences in feed intake, intake patterns, and the composition and functioning of the rumen microbiota. To address this issue, here we set out to establish specific equations that predict EFA intake for an individual cow based on the difference (i.e. the residuals) between observed EFA intake and the predicted EFA intake based on our group-level equations. We studied a database of individual dairy cows (26 experiments; 503 datapoints from three research teams) and we predicted the residuals from 1) dietary and animal-related factors (i.e. full predictions) and 2) animal-related factors only (i.e. field predictions), which are considered more field-amenable. The variance of predicted LA and log ALA intake was explained to 68% by observed LA intake and 66% by observed log ALA intake, respectively. The residuals of LA intake were predicted by dietary ALA content, total FA intake, BW, milk yield and fat content in full predictions, and by BW, feeding level, milk yield and fat content, and sum of milk C4:0 to C14:0 FA in field predictions. The log residuals of ALA intake were predicted by dietary NDF and total FA contents, NDF intake, BW, milk protein, LA and ALA contents, and fat yield in full predictions, and by BW, DM intake (DMI), milk LA and ALA contents, and fat yield in field predictions. The field predictions showed a mo