Subtypes of Alzheimer disease (AD) can be approached from different wa
ys. An intuitive and commonly proposed subtype relates to the age at o
nset of disease. Research on age at onset as a subtype has yielded con
flicting results that may, in part, reflect different measurement appr
oaches. Refinement of the measurement of change in AD should yield a c
learer picture of subtypes, as would closer ties between neuropatholog
ic and clinical work. Thus, it is suggested that future research shoul
d use measurement techniques that more closely characterize AD than do
current approaches. In addition, progress would likely be through att
empts to link clinical findings regarding rates of change to neuropath
ologic findings.