Correlates of cognitive change.

Although there has been considerable interest in identifying potential correlates of cognitive change, results of past studies have been inconsistent. The present study incorporated a number of methodological features intended to maximize sensitivity to detect characteristics of individuals with different amounts of cognitive change. Cognitive change in 5 cognitive abilities was analyzed with 2nd-order latent growth curve models applied to data from a moderately large sample of healthy adults ranging from 18 to 99 years of age (Ns of 4,802 with 1 occasion, 2,265 with 2 occasions, and 1,128 with 3 occasions). There was significant individual difference variance in the longitudinal changes in several cognitive abilities, even in separate analyses of individuals between 18 years of age and 39, between 40 and 64, and 65 and over. Potential correlates of change included measures of self-rated health, vision, mood, personality, and lifestyle. Most of the potential correlates of change had high reliability, and several analyses were based on even more reliable factors determined by the variance common to multiple measures. Despite favorable conditions for detecting correlates of change, there was little evidence that cognitive change was moderated by any of the variables examined. Possible reasons for the inconsistent results regarding correlates of cognitive change are discussed.