Nowadays, I do not believe we have data about the "6 sets" separately. There is no reason to build up studies going like this if you are thinking in dementia appropriately diagnosed in clinical pratice only. Besides, the disease's progressing could be quite different among the subtipes, with early and late onset dementia having very distinct facets.
But you should try looking for "stages in dementia prevalence", in the link nia.nih.gov.
Thank you for your answers! unfortunately did not answer my ask. I know the criteria and prevalence studies but I would like compare with DSM-IV criteria. Well, maybe i'm late because DSM-5 is here... hehehe