OBJECTIVE:

To examine whether occupational complexity of working with data or people, and cognitive or social leisure activity at midlife predicted cognition in advanced old age.

METHODS:

We used 810 eligible participants from Longitudinal Study of Living Conditions of the Oldest Old, a Swedish nationally representative study of individuals aged 77+ with cognitive assessments (an abridged version of the Mini-Mental State Exam) administered in 1992 and 2002 and linked to information about their midlife occupation and leisure activities collected in 1968 and 1981. A bootstrapping technique was applied to examine the direct and interactive associations of occupational complexity and leisure activity with late-life cognition.

RESULTS:

Controlling for demographic and health-related factors from childhood, midlife, and late life, we found that greater work complexity, both with people and with data, and greater participation in cognitive or social leisure activities independently related to better late-life cognitive scores. The complexity-cognition link was moderated by leisure activity such that the cognitive benefit related to the complexity of work-especially complexity of working with people-was rendered insignificant when participation in leisure activities-especially social activities-was above average.

DISCUSSION:

Results are discussed in terms of using work complexity to compensate for lack of leisure activity as well as in terms of promoting leisure engagement to compensate for long-term cognitive disadvantage imposed by working in less challenging occupations.