RISK FOR ALZHEIMER'S DISEASE AND COGNITIVE DECLINE IN PROJECT TALENT ABSTRACT [rev. June 15, 2016] Higher educational attainment is associated with preserved cognitive performance in older age and reduced risk for Alzheimer's Disease (AD) and other dementias. The mechanisms underlying these associations are unclear, but possible explanations include: direct protective effects of more education; preserved capacity or enhanced functioning due to greater cognitive complexity of occupational and leisure activities among people with more education; indirect effects via individual, family, or community-level correlates of educational level; and genetic factors that contribute to both educational level and later-life cognitive functioning. Distinguishing among these alternatives is critical for informing intervention programs to prevent and delay the development of AD and to remediate age-related cognitive decline. These mechanisms will be evaluated using the unique twin/sibling/schoolmate design of Project Talent (PT), a longitudinal study begun in 1960 with a U.S. representative sample of 377,000 high school students who had detailed measures of cognitive abilities, family and school characteristics, educational aspirations and vocational interests. The sample includes 86,000 siblings in 42,000 families, including 2,300 sets of twins and triplets. The twins have been recently located and surveyed (R01-AG043656, Pl: Prescott; Co-Pl: Lyter). Proposed aims are to: (1) locate Project Talent participants in a new follow-up sample of 2500 sibling sets; (2) assess dementia outcomes and cognitive decline in the previously-recruited twin/sibling sample and in the new sample of siblings; (3) combine this outcome information with data collected since 1960 to evaluate the association between educational attainment and risk for dementia and later-life cognitive decline, accounting for cognitive abilities assessed in adolescence; (4) evaluate alternative mechanisms for the association of education with dementia and cognitive decline, including: (a) direct protective effects of higher education: (b) protective effects of cognitive activities; (c) indirect association via early-life cognitive abilities; and (d) indirect association via genetic, family and/or macroenvironmental factors; and (5) document and archive the twin and sibling data for use by other researchers studying the antecedents of cognitive decline. Participants will be assessed with a contemporary battery of cognitive measures harmonized with several of the original 1960 PT measures, allowing direct measurement of change across 1960 to 2017 on multiple cognitive domains; use of adaptive testing to reduce participant burden; and assessment via mailed tablet computers to facilitate measurement of memory and visuospatial abilities. This unique merger of within-family, between-family, within-school and between-school designs controls for genetic and environmental factors that are confounded in other cohort studies and provides an unprecedented opportunity to address causal hypotheses about the mechanisms underlying individual differences in risk for AD, other dementias and cognitive decline. RISK FOR ALZHEIMER'S DISEASE AND COGNITIVE DECLINE IN PROJECT TALENT SPECIFIC AIMS [rev. June 15, 2016] Higher educational attainment is consistently associated with preserved cognition in older age (Albert et al., 1995; Wilson et al., 2009), including reduced risk for Alzheimer's Disease (AD) and other dementias (Bennett et al., 2003; Sharp & Gatz, 2011 ). The mechanisms underlying these associations are unclear, but possible explanations include: direct protective effects of more education; preserved capacity or enhanced functioning due to greater cognitive complexity of occupational and leisure activities among people with more education; indirect effects via individual, family, or community-level correlates of educational level; and genetic factors that contribute to both educational level and later-life cognitive functioning. We will evaluate these alternative hypotheses using the unique twin/sibling/schoolmate design embedded within Project Talent (PT), a longitudinal study begun in 1960 with a U.S. representative sample of 377,000 high school students. The 1960 assessment included extensive measures of cognitive abilities, educational aspirations, vocational interests, and family school and neighborhood characteristics. The sample included 86,000 siblings in 42,000 families, including 2300 sets of twins. Participants are now aged 69-74. As part of R01-AG043656 {Pl: Prescott; Co-Pl: Achorn), we located 96% of PT twins and siblings of twins (including 25% who are deceased). More than 2400 (60% of living) returned surveys on demographics, health history and educational and occupational outcomes for themselves and their siblings. We thus have outcome information on >70% of this sample. In the proposed project, we will assess cognitive functioning and AD in the twin/sib sample and in a parallel sample of siblings drawn from the same high schools as the twins. Including schoolmate controls provides leverage for separating the etiological relevance of family environmental, school and socio- environmental factors on AD risk and cognitive outcomes in later life and evaluating the mediating and moderating roles of educational experiences, occupational activities, and later-life cognitive engagement.
Aim 1. Locate Project Talent participants in a new follow-up sample of 2500 sibling sets. Participants will be sampled using age, gender and ethnicity to adjust for differences in mortality and anticipated participation rates. They will be located using our highly effective tracking methods.
Aim 2. Assess dementia outcomes and cognitive decline in the previously collected twin/sibling sample and in the new sample of sibling sets. The assessment battery will include cognitive measures that are adaptive in format, allowing us to administer tests of multiple abilities from the original (1960) PT assessment and contemporary neuropsychological measures that will allow us to characterize performance across multiple domains, including memory, attention, visuospatial and verbal reasoning. Assessments will use multiple modalities to balance efficiency and account for participant experience. Individuals familiar with computers and having internet access (-70%) will receive a web-based assessment; others will be mailed tablet computers (20%), or receive telephone interviews (10%). Cognitive status of non-participating individuals will be assessed via sibling reports and (for those deceased) matching to Medicare records.
Aim 3. Data collected for Aim 2 will be combined with data collected since 1960 to evaluate the association between educational attainment and risk for dementia and later-life cognitive decline, accounting for cognitive abilities assessed in adolescence. We hypothesize: (3a) The relation between attainment and cognitive decline will be stronger in some cognitive domains (cystallized, Ge) than others (fluid Gf, visuospatial Gv, and memory Gm). (3b) The complexity of mid- to later-life cognitive demands from occupational and leisure activities will be associated with less cognitive decline and lower dementia risk, even after adjusting for health status and physical activity. (3c) Cognitive complexity will partially mediate the association between educational attainment and cognitive status.
Aim 4. The unique twin/sibling/schoolmate design of this study will be used to evaluate alternative mechanisms for the association of education with dementia and cognitive decline, including: (4a) direct protective effects of higher education: (4b) protective effects of cognitive activities; (4c) indirect association via early-life cognitive abilities; and (4d) indirect association via genetic, family and/or macroenvironmental factors.
Aim 5. We will document and archive the twin and sibling data for use by other researchers studying the antecedents of cognitive decline. The 1960 PT data are archived and available from NACDA but the family and twin linking information has not yet been released. We will combine the early and new data, and create user-friendly, anonymous and non-identifiable public use files for dissemination at ICPSR. In summary, this merger of within-family, between-family/within school and between-school designs controls for genetic and environmental factors that are confounded in other cohort studies. PT's sample size, prospective design, measurement of multiple cognitive abilities in adolescence and older adulthood, combined with the twin/sibling/schoolmate structure provide a unique opportunity to shed light on the mechanisms underlying individual differences in risk for AD, other dementias and cognitive decline.
