Age, vascular risk, and cognitive decline in an older, British, African-Caribbean population

To investigate associations between baseline factors and subsequent cognitive decline in an older African-Caribbean population.

Three-year follow-up of a community sample.

Registration lists for seven primary care services in south London, United Kingdom.

Of 290 Caribbean-born people aged 55 to 75 recruited at baseline, 216 (75%) were reinterviewed after a 3-year period. Sufficient data for analysis were present for 207.

Tests of verbal memory (immediate and delayed), orientation, and attention were administered at baseline and at 3 years. A composite measure of cognitive change was derived through factor analysis. Participants in the lowest quintile were defined as having cognitive decline and were compared with the rest of the sample with respect to baseline characteristics.

Cognitive decline was strongly associated with increased age, but not with sex, education, or occupation. It was also not directly associated with hypertension, diabetes mellitus, or exercise, but the age-decline association was significantly stronger in people with diabetes mellitus and was lower in those reporting vigorous physical exercise at baseline. When cognitive decline was investigated separately for individual tests, both these patterns of association were particularly strong for decline in delayed verbal recall.

The association between increased age and cognitive decline, particularly decline in verbal recall, was stronger in people with diabetes mellitus (but not hypertension) and weaker in people who were more physically active.