The ultimate goal of our project, summarized in our Aim 1, is to examine the long term effects of lifetime cognitive engagement, the short term effects of cognitive engagement in old age, and the mediating effect of lifetime cognitive engagement on those short term effects. This examination is made possible by the large and rich data we will use, innovative use of this data to create measures of cognitive engagement, and a statistical model that can jointly estimate long term associations and short term effects. These estimates will help assess the importance of the long term effects of cognitive engagement over the life course through the mechanism of cognitive reserve and the shorter term effects of cognitive engagement at older ages, through ?use-it-or-lose-it?.
This project will develop measures of cognitive engagement, defined as the extent to which people use their brains, though the lives of a large set of subjects, and assess the dynamic and interactive role of cognitive engagement in young and middle age and cognitive engagement in old age in cognitive decline and the onset and patterns of dementia. We will use rich data from the large, longitudinal but complex Health and Retirement Study (HRS), including its supplements, to construct these measures and carry out various analyses with them. The goal of our Aim 2 is to create measures of lifetime cognitive engagement using complete and harmonized histories of education and employment and detailed descriptions of jobs, including requirements of cognitive and social tasks. In our Aim 3, we will create measures of cognitive engagement using many variables on work and non-work activities, scattered around different parts of the survey and available for different subsets of the respondents. Our Aim 4 is to complement the activity-based measures by a novel measure based on response patterns to cognitively demanding survey questions, such as don?t know responses, rounded and crude value responses, longitudinal variation of responses, and other aspects of response style.