Campbell and colleagues noted that the LONGSCAN study had limitations that extend to the current analysis, including lack of data on intimate partner violence in participating households, substance abuse, and other factors also known to affect risks of child abuse.