b'AGING IN THE 21ST CENTURYimpairment without dementia (CIND), and nor-mal cognitive functioning. These data can nowDementia Prevalencebe used for validation studies and to developAs the US population becomes older predicted probabilities of dementia and CIND in the HRS (Langa et al. 2005).in the decades ahead, we can expect a One report demonstrates how diagnosticlarge growth in the number of people data from ADAMS can be used to develop methods for classifying HRS respondents inwith AD and dementia, simply because the same diagnostic categories on the basis ofthe risk for these conditions is so much cognitive measures available in the larger sur- higher as we age. Some have called this vey (Crimmins et al. 2011). Overall, the cogni-tive tests in the HRS predict dementia cases ina looming dementia epidemic. Figure 74% of the sample who could answer questions2-1 shows higher rates of dementia in for themselves. Not surprisingly, classification of CIND was not as high.older age groups. For example, in 2012, dementia affects 3% of those aged 65 to 74 compared to 30% for those aged 85 and older.FIGURE 2-1Trends in dementia prevalence by age: 1992-2012 Through a combination of factors like Source: HRS 1998-2012. education and better medical treatment for conditions that affect dementia risk, 35%rates of dementia are slowly trending 30%25% downward, as Figure 2-1 also shows.20% In the youngest group, those aged 6515% to 74, the percentage with dementia 10% moves steadily from 5% in 1998 to 3% in 5% 2012. The next oldest group, those aged 0% 1998 2000 2002 2004 2006 2008 2010 2012 75 to 84, experienced even larger de-Demented aged: 65-74 75-84 85+ clines, from 14% in 1998 to 10% in 2012.38'