Assessing Dementia Prevalence in the Wisconsin Longitudinal Study: Cohort Profile, Protocol, and Preliminary Findings

Abstract

Background: There is growing consensus that non-genetic determinants of dementia can be linked to various risk- and resiliency-enhancing factors accumulating throughout the lifespan, including socioeconomic conditions, early life experiences, educational attainment, lifestyle behaviors, and physical/mental health. Yet, the causal impact of these diverse factors on dementia risk remain poorly understood due to few longitudinal studies prospectively characterizing these influences across the lifespan.

Objective: The Initial Lifespan's Impact on Alzheimer's Disease and Related Dementia (ILIAD) study aims to characterize dementia prevalence in the Wisconsin Longitudinal Study (WLS), a 60-year longitudinal study documenting life course trajectories of educational, family, occupational, psychological, cognitive, and health measures.

Methods: Participants are surveyed using the modified Telephone Interview for Cognitive Status (TICS-m) to identify dementia risk. Those scoring below cutoff undergo home-based neuropsychological, physical/neurological, and functional assessments. Dementia diagnosis is determined by consensus panel and merged with existing WLS data for combined analysis.

Results: Preliminary findings demonstrate the initial success of the ILIAD protocol in detecting dementia prevalence in the WLS. Increasing age, hearing issues, lower IQ, male sex, APOE4 positivity, and a steeper annualized rate of memory decline assessed in the prior two study waves, all increased likelihood of falling below the TICS-m cutoff for dementia risk. TICS-m scores significantly correlated with standard neuropsychological performance and functional outcomes.

Conclusion: We provide an overview of the WLS study, describe existing key lifespan variables relevant to studies of dementia and cognitive aging, detail the current WLS-ILIAD study protocol, and provide a first glimpse of preliminary study findings.

Keywords: Alzheimer’s disease; dementia; epidemiologic determinants; health risk behaviors; prevalence.

Fig. 1.

WLS graduate and sibling sample retention and attrition rates for each of the prior study waves. For graduates, the grayed boxes for Wave 2 reflect response metrics from a parent-administered survey, and do not capture actual graduate participation. Percent retention rates are calculated for all eligible participants for each study wave excluding known deceased participants. Overall retention rates for the graduate sample are quite high across study waves (74–90%), while sibling sample retention are a bit lower (45–59%). A full characterization of WLS retention data is publicly available at: https://www.ssc.wisc.edu/wlsresearch/documentation/retention/cor1004_retention.pdf.

Fig. 2.

Overview of Wisconsin Longitudinal Study (WLS) – Initial Lifetime Impact on Alzheimer’s Disease and Related Dementia (ILIAD) study protocol. The general research strategy involves an initial phone-based interview of the active and living WLS cohort through partnership with the University of Wisconsin Survey Center to identify individuals at risk for dementia or cognitive impairment, and to obtain updated health and social measures collected in prior study waves. Those participants selected as at-risk for dementia based on telephone-based cognitive screening undergo an intensive follow-up in-home neuropsychological and medical assessment. Data gathered during the in-home comprehensive assessment is reviewed by a multidisciplinary panel (physician, clinical neuropsychologist, and nurse practitioner) to arrive at a consensus clinical diagnosis of dementia. For cases in which the consensus panel is unable to clearly establish a diagnosis based on in-home assessment data and medical records, participants are invited for an on-site assessment at the University of Wisconsin (UW) Madison, Alzheimer’s Disease Research Center (ADRC) to obtain clinical dementia biomarker data (MRI, blood draw) and/or additional neuropsychological assessment to support a more definitive diagnosis.

Fig. 3.

A) Participants selected for Phase 2 (red line: TiCS-m ≤ 28) scored significantly lower than those not selected (blue line: TICS-m > 28) on delayed recall for a 10-item word list assessed at the prior two study waves (2004 and 2011). B) Those selected for Phase 2 (red) also showed a steeper annualized rate of change on delayed word recall between the two timepoints, compared to cases not selected (blue).