Technology Use Among Caregivers for Persons With and Without Cognitive Impairment

Abstract This study aims at describing technology use among caregivers for middle-aged or older adults with and without cognitive impairment (CI) and examining whether the associations between technology use, caregiver strain, and social support differ by care-recipient CI status. Online data from caregivers (n=561) for adults at aged 50 years and older were analyzed from a national caregiver and technology survey. Multiple binary items were used to indicate caregivers’ use of various devices (smartphone, computer, e-reader, and wearable activity tracker) and applications (communication, online banking, navigation, online entertainment, medication alert/tracker, and physical activity tracker). Predictors were care-recipient CI status (having been diagnosed with cognitive problems versus no cognitive problems), caregiver strain, and social support. Multivariable logistic regression models were fitted for each technology and for testing effect modification by care- recipient CI status. All models were adjusted for total caregiving hours and caregiver age, sex, race/ethnicity, financial status, and residence. Almost half (47%) reported their care-recipient was diagnosed with CI. Caregivers for those with CI were more likely to use e-readers (adjusted odds ratio (AOR) = 1.55, p=.040), wearable activity trackers (AOR=1.77, p=.013), and medication alerts/trackers (AOR=2.59, p<.001). Generally, greater caregiving strain and social support were positively associated with use of multiple technologies (p<.05). No effect modification of caregiving strain and social support by CI status was observed (p>.05). Technology use differences among caregivers of persons with CI may be driven by care recipients’ unique situations and demands. Future research should identify technology use benefits on caregiver health-related quality of life.

the precursor to Alzheimer's disease and related dementias. Recent developments using online automated monitoring of financial transactions promise a new way to identify the earliest signs of cognitive decline. We examine the feasibility of using secure online technology to link ongoing financial activity monitoring data with other objective measures of function and cognition in a cohort of independent living older adults. To date, 73 older adults (mean age = 76.8, MoCA = 25.9) have enrolled and are participating in a 12-month online financial monitoring program that tracks account activity and generates alerts for unusual or irregular transactions. At baseline participants are administered a battery of neuropsychological tests and the Financial Capacity Instrument (FCI), a measure of financial capacity using tasks of everyday financial activity. Financial monitoring data are collected continuously, and are summarized and reported monthly. Younger participants had more online transactions and higher FCI scores. FCI total score was positively correlated with animal fluency (p < .02), Trails A (p < .03) and B (p < .0001), and visual memory (p < .0008). Number of online transactions in one month was correlated with FCI score, and Trails B (faster time to completion). Lower MoCA scores were associated with higher number of alerts per month.

CONSERVATION OF RESOURCES THEORY: TECHNOLOGY AND CAREGIVER STRAIN
Shinduk Lee, 1 Marcia Ory, 1 Deborah Vollmer Dahlke, 2 Tiffany Shubert, 3 Steve Popovich, 4 and Matthew Smith, 1 1. Texas A&M University,College Station,Texas,United States,2. DVD Associates,LLC,Austin,Texas,United States,3. Shubert Consulting,Chapel Hill,North Carolina,United States,4. Clairvoyant Network,Austin,Texas,United States Using the Conservation of Resources Theory, this study examined how caregiver strain was influenced by care recipients' use of falls alert wearables. Online survey data from 486 unpaid caregivers for adults aged 50 and older were analyzed. Structural equation modeling was used to test the following hypotheses: (1) caregivers with fewer financial resources would engage in fewer resource conservation strategies (e.g., care recipients' use of falls alert wearables); (2) resource conservation strategy engagement would be associated less resource loss; and (3) the effect of resource conservation strategies on caregiver strain would be less salient than the effect of resources used on caregiving (e.g., time and social support). The hypothesized model had a good model fit (CFI=.910), with SRMR (.060) and RMSEA (.062) being close to .05. All hypothesized paths were statisticallysignificant, except for the direct effect of using falls alert wearables on social support (p=.076) and caregiver strain (p=.135). As hypothesized, higher income was associated with greater likelihood of using falls alert wearables (b=.10, p>.022). Technology use was associated with less time spent on caregiving (b=-.16, p<.001) and had statistically-significant indirect effects on caregiver strain (b=-.03, p=.008). The total effect of using falls alert wearables (b=.04, p=.394) on caregiver strain was less powerful than the effect of time (b=.20, p<.001) or social support (b=-.28, p<.001). Study findings suggest the benefits of using falls alert wearables to alleviate time-related burdens and downstream caregiver strain among unpaid caregivers. Future efforts should investigate the relative advantage of wearables for other caregiving purposes.

I DO HELP: OLDER ADULTS AS DIGITAL MEDIA SUPPORT PROVIDERS FOR THEIR PEERS
Amanda Hunsaker, Minh Hao Nguyen, Jaelle Fuchs, Gökçe Karaoglu, Teodora Djukaric, and Eszter Hargittai, University of Zurich, Zurich, Switzerland Older adults include highly sophisticated digital media users among their numbers for whom diverse methods of using online technology are a daily occurrence. Given that some older adults are quite tech-savvy, it follows that they may also provide digital media support to others. This study examines technological support-giving abilities and experiences of older adults. We completed in-depth qualitative interviews with older adults (ages 59+) in Hungary, the Netherlands, Switzerland, and Turkey (N=63) exploring: (1) whether older adults give digital media support and reasons for (not) doing so; (2) how older adults provide support and to whom; and (3) experiences of providing support with IT security specifically. The research team conducted consensus meetings to identify themes and coding schemes, and memo writing to share findings and ensure reliability across and within coders. Overall, a large number of our participants reported giving digital media support, primarily within their social circles. The kinds of help provided highly varied from navigating the Internet to choosing devices and managing security settings. We also find mutual support -support given to each other -an important domain for how people in this age group share assistance for technical problems. While some participants did not provide help with digital technology because they lacked confidence to do so, others believed they could, but were never asked. That such a prominent amount of support providers exist in this age group implies that peerled technical support approaches may be especially salient and effective in helping older adults use digital media.

TECHNOLOGY USE AMONG CAREGIVERS FOR PERSONS WITH AND WITHOUT COGNITIVE IMPAIRMENT
Shinduk Lee, 1 Matthew Smith, 1 Deborah Vollmer Dahlke, 2 Tiffany Shubert, 3 Steve Popovich, 4 and Marcia Ory, 1 1. Texas A&M University,College Station,Texas,United States,2. DVD Associates,LLC,Austin,Texas,United States,3. Shubert Consulting,Chapel Hill,North Carolina,United States,4. Clairvoyant Network,Austin,Texas,United States This study aims at describing technology use among caregivers for middle-aged or older adults with and without cognitive impairment (CI) and examining whether the associations between technology use, caregiver strain, and social support differ by care-recipient CI status. Online data from caregivers (n=561) for adults at aged 50 years and older were analyzed from a national caregiver and technology survey. Multiple binary items were used to indicate caregivers' use of various devices (smartphone, computer, e-reader, and wearable activity tracker) and applications (communication, online banking, navigation, online entertainment, medication alert/tracker, and physical activity tracker). Predictors were care-recipient CI status (having been diagnosed with cognitive problems versus no cognitive problems), caregiver strain, and social support. Multivariable logistic regression models were fitted for each technology and for testing effect modification by care-recipient CI status. All models were adjusted for total caregiving hours and caregiver age, sex, race/ethnicity, financial status, and residence. Almost half (47%) reported their care-recipient was diagnosed with CI. Caregivers for those with CI were more likely to use e-readers (adjusted odds ratio (AOR) = 1.55, p=.040), wearable activity trackers (AOR=1.77, p=.013), and medication alerts/trackers (AOR=2.59, p<.001). Generally, greater caregiving strain and social support were positively associated with use of multiple technologies (p<.05). No effect modification of caregiving strain and social support by CI status was observed (p>.05). Technology use differences among caregivers of persons with CI may be driven by care recipients' unique situations and demands. Future research should identify technology use benefits on caregiver health-related quality of life.

ARE DIFFERENCES IN BEDTIMES BAD FOR RELATIONSHIPS? ASSOCIATIONS WITH ATTACHMENT AND CONFLICT IN MIDDLE-AGED COUPLES Joshua Novak, 1 and Stephanie Wilson, 2 1. Auburn University, Auburn, Alabama, United States, 2. Southern Methodist University, Dallas, Texas, United States
A robust body of literature has found birdirectional associations between sleep quality and marital quality in couple relationships (Hasler & Troxel, 2010;Pearlin, 2010). Additionally, dyadic research shows that differences in couples' bed time routines and habits is associated with mental health outcomes (Chen, 2018), however the literature has not connected them with other marital processes that are mutable and clinically relevant. Attachment theory provides a clinically relevant framework that captures both interpersonal marital processes such as relationship conflict as well intrapersonal processes of individual emotional safety-essentially individuals' personal strategies to balance closeness and distance in a relationship (Feeney, 2002;Rhodes et al., 2001). The two main attachment styles related to sleep processes are attachment avoidance and attachment anxiety (Collins et al., 2002;Gun, 2015;Troxel, 2007). Utilizing data from 234 couple dyads, we investigated if differences in partners' bed times is associated with conflict frequency and attachment avoidance using a structural equation modeling approach. We controlled for a number of important factors and tested our hypothesized model against two plausible alternative models. Results revealed that greater difference in partners' bed times was associated with higher conflict frequency for both husbands and wives through higher men's attachment avoidance. Our findings highlight previous research on matched vs. unmatched couples on sleep routines, habits, and chronotypes (both morning or night vs. different; Larson et al., 1991) but highlight mutable and clinically relevant constructs for intervention. Implications for health promotion and marital therapy will be discussed.

ASSISTIVE RELAXATION THERAPY FOR OLDER ADULTS WITH INSOMNIA AND MILD COGNITIVE IMPAIRMENT: A PILOT STUDY
Miranda McPhillips, 1 Junxin Li, 2 E. John Ward III, 3 and Nalaka Gooneratne, 1 1. University of Pennsylvania,Philadelphia,Pennsylvania,United States,2. Johns Hopkins School of Nursing,Baltimore,Maryland,United States,3. University of Pennsylvania,School of Medicine,Philadelphia,United States Insomnia symptoms are prevalent in older adults with mild cognitive impairment (MCI) and can pose treatment challenges. Our objective was to test the preliminary efficacy of tablet-based assistive relaxation therapy (ART) to improve insomnia symptoms in community-dwelling older adults with MCI. ART involves breath-based relaxation techniques coupled with a physical anchoring task to redirect thoughts and disengage from pre-sleep anxiety-provoking cognitions. Using a pilot randomized controlled non-crossover design, 20 participants recruited from one urban adult day center were allocated in a 1:1 ratio to intervention or education only control group for a treatment period of two weeks. Our final sample (n=20) was balanced on all demographic and clinical variables and consisted of Black (100%), female (75%), older adults (mean age 68.85 ± 7.29) with mean Montreal Cognitive Assessment scores of 21.2 ± 2.48. All participants at baseline had insomnia symptoms (mean Insomnia Severity Index (ISI) score 15.8 ± 3.78) and poor sleep quality (mean Pittsburgh Sleep Quality Index (PSQI) score 12.95 ± 0.70); half had daytime sleepiness (Epworth Sleepiness Scale (ESS) score 10.15 ± 1.07). Compared to baseline, participants improved on ISI (9.83 ± 1.32; p=0.0002), PSQI (9.11 ± 1.02; p=0.0016) and ESS (8.17 ± 0.86; p=0.08). The intervention group had statistically significant mean change scores on ISI compared to the control (-7.5 ± 1.37 vs. -3.88 ± 1.48; p=.0461). There were no statistically significant between group differences on PSQI or ESS. Our preliminary results suggest ART therapy is an effective treatment for insomnia symptoms in older adults with MCI. Few studies have longitudinally investigated the association between objectively measured sleep and time to develop dementia. This study leverages polysomnography (PSG) sleep data extracted from the VA national electronic health records (VA-EHR) to assess the association between sleep and time to develop dementia. We identified 61,165 PSG reports from the VA-EHR from 2000 to 2019 using CPT codes. Patients