Using the minimum data set to predict weight loss in nursing home residents.

In this longitudinal exploratory study, items from the minimum data set (MDS) were used to predict resident weight changes in older adults (N = 77) from three nursing homes. At the end of the study period, three variables from the last available MDS were significant predictors of weight loss: 1) leaves 25% or more of food uneaten at most meals, 2) receiving an antianxiety medication, and 3) not having the ability to feed oneself independently.