Response to comment on “A proteomic surrogate for cardiovascular outcomes that is sensitive to multiple mechanisms of change in risk”