H. Christensen et al., Dispersion in cognitive ability as a function of age: A longitudinal studyof an elderly community sample, AGING NEURO, 6(3), 1999, pp. 214-228
This longitudinal study investigated whether age is associated with increas
ed dispersion among major domains of cognitive ability. Three samples were
examined: the full sample of 760 elderly community dwellers aged 70 years a
nd older who were tested in 1990; a subset of the original sample who died
between testing occasions; and the sample of 426 who survived with full dat
a sets in 1994 (followed up for a mean 3.5 years). Dispersion, as measured
by the within-individual standard deviation of ability scores and by the wi
thin-individual deviations from crystallized intelligence for speed, memory
and spatial functioning, was significantly correlated with age in all thre
e samples at Wave 1 and at Wave 2 (for the longitudinal sample). The rate a
t which dispersion increased was not significantly correlated with age. In
a more derailed analysis of the 426 survivors, dispersion as a function of
age was similar for demented persons within this sample. those without deme
ntia, those with poor and excellent educational levels, and those with a ph
ysical disability. Activities of daily living was a predictor of larger-tha
n-average changes in dispersion - but not age, education, or activity. Grea
ter dispersion was associated with faster deterioration in memory and speed
performance. Contrary to some recent reports, there was evidence for great
er within individual variability among cognitive domains in older individua
ls.