Older adults living with Alzheimer's disease (AD) require progressively greater levels of support (e.g., in-home assistance and supportive facilities) 1, 2. Yet, older adults and their families vary widely in how they perceive and plan for future support needs 3. Recognizing these diverse attitudes toward long-term care (LTC) planning, we aimed to characterize and model the distinct phenotypes that emerge as older adults consider their future AD-related support needs. The Plan Your Lifespan study examines older adult decision-making processes about support, in the event of developing AD 4. Subjects were recruited from the LitCog study; a cohort of older adults age 65+ who participate in extensive neuropsychological cognitive testing every 2.5 years 5. Subjects were provided PlanYourLifespan.org, which facilitates aging-in-place and LTC decision-making 6-9. Subjects were interviewed every 6 months for 42-months, about cognitive, social, functional, environmental factors, and hypothetical scenarios such as if they were to develop AD, have they decided on support preferences. All interviews from baseline-42 months were analyzed. We employed mixed analytic methods through qualitative coding and generalized linear mixed modeling (GLMM). Mixed-methods analysis included analyzing open-ended, qualitative responses about perceptions of planning in the event of developing AD. The responses were coded using constant comparative analysis with triangulation of themes. GLMM was used to identify factors associated with statistically significant increases or decreases in the likelihood of planning for AD. Of the 293 subjects enrolled (mean age = 73.5 years, 72.7% female, 40.4% under-represented minority, 40.6% limited health literacy), almost half (47.4%, n = 139) self-reported worsening memory loss at 18 months post-baseline, with 42.4% (n = 59) experiencing memory loss episodes weekly and 18% (n = 25) monthly (Table 1). Neuropsychological testing revealed 22.2% (n = 65) had mild or moderate cognitive impairment. Nearly half of those subjects (46.2%, n = 30) denied experiencing any cognitive loss. Of subjects testing at normal cognitive levels, 47.8% (n = 104) reported new or worsening memory loss. At baseline, 15% (n = 44) had never considered the possibility of developing AD in the future. Qualitative findings revealed four planning phenotypes: (a) Deniers, (b) Do-gooders, (c) Dumpers, and (d) Defeated (Figure 1). Deniers rejected the need to plan for future health events, such as AD. They displayed decreased perceived support needs (OR 0.357; p < 0.05), higher working memory (OR 0.88; p < 0.05), higher inductive reasoning (OR 0.90; p < 0.05), higher processing speed (OR 0.95; p < 0.05), and higher busyness (OR 0.95; p < 0.05). Do-gooders created LTC plans to avoid burdening others. Do-gooders exhibited adequate health literacy (OR 5.06; p < 0.05), increased self-efficacy (OR 1.07; p < 0.01), and more cognitive impairment (OR 1.46; p < 0.05). They reported sufficient social support (OR 2.82; p < 0.05), prior experiences with others with AD or memory loss (OR 1.90; p < 0.05), completion of a living will (OR 2.43; p < 0.05), and had a higher number of chronic conditions (OR 1.20; p < 0.05). Dumpers left their future health care planning and assist with support needs to others. This dumping was possible since they had sufficient social support (OR 3.39; p < 0.05) and adequate social networks (OR 1.08; p < 0.05). Defeated subjects did not plan due to perceived internal (e.g., too sick) or external limitations (e.g., lack of savings). They exhibited below average health activation scores (OR 0.65; p < 0.05), limited health literacy (OR 1.84; p < 0.05), decreased physical function (OR 0.95; p < 0.05), increased anxiety scores (OR 1.04; p < 0.05), and decreased self-efficacy (OR 0.96; p < 0.05). When planning for future LTC support needs, four distinct phenotypes for older adults were identified. Deniers believed that they would never require future support and avoided future-oriented thinking. Clinicians might be able to intercede in realistically pointing out their prognosis and future needs. Do-gooders were highly motivated to avoid burdening others. With prior experiences with others who had AD, do-gooders may have a greater understanding of the effect AD has on the family. Do-gooders commonly tested at levels of cognitive decline, and their awareness of cognitive decline and prior experience with others with AD could explain the urgency of creating LTC plans 9. Dumpers had sufficient social support and social networks on which they could depend. Prior research has shown that while some older adults anticipate their assistance, families may not be aware of these designs 10. Therefore, providers can help encourage communication between families and older adults. Defeated subjects had decreased self-efficacy, increased anxiety, and often unreliable social networks. They were overwhelmed and unable to form plans. Social work might be tapped to intervene with these older adults. Across all phenotypes, planning behaviors were dependent on health literacy levels, presence and availability of social support, and personal experience with others who had AD. Future research will reveal how these phenotypes evolve over time, with changing social networks, personal experiences, and health needs. Identifying these phenotypes and trajectories will be essential for designing interventions that assist older adults at risk for AD in making timely and informed long-term care decisions. All authors met criteria for authorship by (1) conception and design of the study: L.A.L., V.R.-Z., (2) data acquisition: APM, AS, AM, R.R.R., V.R.-Z., (3) analysis and interpretation of data: All authors. (4) manuscript drafting: All authors. (5) revising the manuscript critically for important intellectual content: All authors. (6) approval of the version of the manuscript to be published: All authors. This research is supported through grants from the NIH/NIA (R01AG058777, R01AG30611, and P30AG059988). All statements in this manuscript, including its findings and conclusions, are solely those of the authors. The authors declare no conflicts of interest.
Miller‐Winder et al. (Fri,) studied this question.