Understanding Attrition and Bolstering Retention in a Longitudinal Panel of Older Adults: ORANJ BOWL

Predictors of Retention in the 1Florida Alzheimer's Disease Research Center (ADRC) Over Two Waves

Background: Attrition is a significant methodological concern in longitudinal studies. Sample loss can limit generalizability and compromise internal validity. Methods: Wave one (n = 346) and wave two follow-ups (n = 196) of the 1Florida ADRC clinical core were examined using a 24-month visit window. Results: The sample (59% Hispanic) demonstrated retention rates of 77.2% and 86.2% in waves one and two, respectively. Predictors of lower retention in wave one included older age, amnestic MCI or dementia, and lower cognition and function scores. Completing a baseline MRI and lack of hippocampal atrophy were associated with higher retention in both waves. In wave two, a greater neighborhood disadvantage score was associated with attrition. Discussion: Predictors of retention changed over time, possibly due to the early withdrawal of the most vulnerable in the initial wave. Understanding predictors of retention can facilitate retention strategies, reduce attrition, and increase the validity of findings.

Why do older adults decline participation in research? Results from two deprescribing clinical trials

BACKGROUND

Heterogenous older adult populations are underrepresented in clinical trials, and their participation is necessary for interventions that directly target them. The purpose of this study was to evaluate reasons why hospitalized older adults declined participation in two deprescribing clinical trials.

METHODS

We report enrollment data from two deprescribing trials, Shed-MEDS (non-Veterans) and VA DROP (Veterans). For both trials, inclusion criteria required participants to be hospitalized, age 50 or older, English-speaking, and taking five or more home medications. Eligible patients were approached for enrollment while hospitalized. When an eligible patient or surrogate declined participation, the reason(s) were recorded and subsequently analyzed inductively to develop themes, and a chi-square test was used for comparison (of themes between Veterans and non-Veterans).

RESULTS

Across both trials, 1226 patients (545 non-Veterans and 681 Veterans) declined enrollment and provided reasons, which were condensed into three themes: (1) feeling overwhelmed by their current health status, (2) lack of interest or mistrust of research, and (3) hesitancy to participate in a deprescribing study. A greater proportion of Veterans expressed a lack of interest or mistrust in research (42% vs 26%, chi-square value = 36.72, p < .001), whereas a greater proportion of non-Veterans expressed feeling overwhelmed by their current health status (54% vs 35%, chi-square value = 42.8 p < 0.001). Across both trials, similar proportion of patients expressed hesitancy to participate in a deprescribing study, with no significant difference between Veterans and non-Veterans (23% and 21%).

CONCLUSIONS

Understanding the reasons older adults decline participation can inform future strategies to engage this multimorbid population.

Improving Retention of Diverse Samples in Longitudinal Research on Developmental Disabilities

Developmental disabilities (DD) research has depended on volunteer and clinical samples, with limited racial/ethnic diversity. This study focused on improving diversity and retention in DD research. The sample included 225 parents with a child with DD and 4,002 parents without children with DD from diverse racial/ethnic groups, drawn from Midlife in the United States, a national longitudinal study. Unexpectedly, parents of children with DD from diverse racial/ethnic groups were more likely to participate longitudinally than other groups. Relative participant payment was a factor that enhanced their likelihood of retention. This research illustrates how large national studies can be leveraged to increase representativeness and ongoing participation of diverse racial/ethnic groups, especially in combination with other factors, such as parenting a child with DD.

Stimulus material selection for the Dutch famous faces test for older adults

Worldwide, approximately 22% of all individuals aged 50 years and older are currently estimated to fall somewhere on the Alzheimer's disease (AD) continuum, which can be roughly divided into preclinical AD, mild cognitive impairment (MCI), and AD dementia. While episodic memory loss (among other aspects) is typically required for a diagnosis of AD dementia, MCI is said to have occurred when cognitive impairment (including memory loss) is worse than expected for the person's age but not enough to be classified as dementia. On the other hand, preclinical AD can currently only be detected using biomarkers; clinical symptoms are not apparent using traditional neuropsychological tests. The main aim of the current paper was to explore the possibility of a test which could distinguish preclinical AD from normal aging. Recent scientific evidence suggests that the Famous Faces Test (FFT) could differentiate preclinical AD from normal aging up to 5 years before a clinical AD diagnosis. Problematic with existing FFTs is the selection of stimulus material. Faces famous in a specific country and a specific decade might not be equally famous for individuals in another country or indeed for people of different ages. The current article describes how famous faces were systematically selected and chosen for the Dutch older (60+) population using five steps. The goal was to design and develop short versions of the FFT for Dutch older adults of equivalent mean difficulty. In future work, these nine parallel versions will be necessary for (a) cross-sectional comparison as well as subsequent longitudinal assessment of cognitively normal and clinical groups and (b) creating personalized norms for the normal aged controls that could be used to compare performance within individuals with clinical diagnoses. The field needs a simple, cognitive test which can distinguish the earliest stages of the dementia continuum from normal aging.

Adverse childhood experiences, adult trauma, and depressive symptom trajectories

OBJECTIVES

Associations among adverse childhood experiences prior to age 18 (ACEs), subjective reports of trauma during Hurricane Sandy, and trajectories of depressive symptoms reported by community-dwelling older people were examined.

METHODS

We analyzed 6 waves of data from 5,688 people aged 50-74 recruited in 2006 and followed for 12-years using multilevel mixed effects models.

RESULTS

We found that: (1) people who experienced ACEs had trajectories of depressive symptoms in late life that were higher than people not having these experiences, (2) people experiencing two or more ACEs were more likely to report fear and distress when Hurricane Sandy hit than people experiencing either one or no adverse childhood experiences, and (3) while both ACE exposure and peri-traumatic stress were associated with trajectories having higher levels of depressive symptoms, the risk associated with ACEs (especially multiple ACEs) was greater.

CONCLUSION

Findings support life course stress theories including the cumulative inequality theory and stress proliferation theory, suggesting that inequalities are manifested over the life course and that people experiencing adversity during childhood are at increased risk of experiencing adversity in late life. By studying the relationship between adverse childhood experiences and response to Hurricane Sandy our findings demonstrate that adverse childhood experiences can alter the way traumatic events in adulthood are experienced. This finding in turn, has important implications for clinical practice, as it identifies a group of people likely to be at risk for adult trauma.

Association of social disengagement with health status and all-cause mortality among community-dwelling older adults: evidence from the Otassha study

This study examined the impact of disengagement on health status and mortality among community-dwelling older adults in Japan. Disengagement from society was operationally defined as dropping out of a longitudinal survey. A follow-up mail survey was conducted, in 2014, among respondents (n = 3696) of the baseline mail survey. Step-by-step follow-up surveys (FLs), including simplified mail, postcard, and home-visit surveys, were sent to participants who did not respond. Disengagement levels were defined according to the response to the FLs as zero (mail survey), low (simplified mail survey), middle (postcard survey), high (home-visit survey), and highest (non-responders to the home-visit survey). After adjusting for health status at baseline, the proportion of respondents self-rated as "not healthy" during FLs was significantly higher in the high-level than in the zero-level group. The proportion of respondents reporting a "once a week or less" frequency of going outdoors during FLs was significantly higher in the low-, middle-, and high-level groups than in the zero-level group. Mortality rates were significantly higher in the high and highest levels than in the zero-level group. Higher disengagement levels increased the risk of lower health status and mortality, suggesting an urgent need to prevent societal disengagement among older adults.

Subjective Successful Aging: Measurement Invariance Across 12 Years

BACKGROUND AND OBJECTIVES

Debates about how to define successful aging have dominated gerontology for over 60 years. Regardless of how successful aging is conceptualized, in order to accurately understand how the construct changes over time and how it differs between people of varying ages, successful aging must be measured with instruments that are valid, reliable, and have measurement invariance. These analyses focus on subjective successful aging and examine the extent to which a reliable, valid, 3-item scale has measurement invariance across 12 years for individuals aged 50-86.

RESEARCH DESIGN AND METHODS

We analyzed 5 waves of data collected from a panel of 5,688 community-dwelling people aged 50-74 when recruited in 2006. We tested measurement invariance using the standard 4 nested steps,, introducing increasing parameter constraints at each step. Analyses were conducted using Mplus 7.

RESULTS

Analyses revealed that the 3-item scale measuring subjective successful aging has adequate measurement invariance across time. We demonstrated that the scale has configural, metric, and scalar invariance by most standard metrics. Only residual invariance was not supported. However, because residuals are not part of the latent factor, invariance of the item residuals is inconsequential to interpretation of latent mean differences.

DISCUSSION AND IMPLICATIONS

Findings provide the foundation needed for researchers to examine change in subjective successful aging over time, differences in subjective successful aging between people of varying ages, and predictors of subjective successful aging, confident that the scale has adequate measurement invariance.

Long-Term Impact of Hurricane Sandy Exposure on Positive and Negative Affect: The Role of Perceived Social Support

OBJECTIVES

Natural disasters can have devastating, long-lasting effects on the mental health of older adults. However, few studies have examined associations among disaster exposure and positive and negative affect, and no longitudinal studies have investigated the extent to which pre-disaster perceived social support affects these associations. These analyses examine the associations among pre-disaster perceived social support, disaster exposure, and positive and negative affect experienced by community-dwelling older adults four years after Hurricane Sandy, controlling for pre-disaster affect.

METHODS

Self-reported data collected before and after Hurricane Sandy from participants (aged 50-74 years) in the ORANJ BOWL panel (N=2442) were analyzed using linear regression models.

RESULTS

Higher levels of peri-traumatic stress experienced during Hurricane Sandy and greater hardship experienced after the storm were associated with more negative affect four years following the disaster. Higher perceived social support at baseline was related to more positive affect and less negative affect both before and after the hurricane. Social support did not moderate the effect of hurricane exposure on either positive or negative affect.

DISCUSSION

Findings suggest that psychological effects may persist years after natural disasters and that more effective interventions may be needed during and after a disaster. While social support is critical to positive and negative affect in general, its buffering effects when disaster strikes may be limited.

Change in Loneliness Experienced by Older Men and Women Living Alone and With Others at the Onset of the COVID-19 Pandemic

Building on theory suggesting that loneliness is distinct from living arrangements, social isolation, and perceived social support, we examined change in loneliness for older people at the onset of the COVID-19 pandemic. Analyzing 14-years of data with multilevel mixed-effects models, we found higher levels of loneliness among people living alone, people more socially isolated, and people with less perceived support. Gender affected changes in loneliness, controlling for social isolation, perceived support, living arrangements, age, education, income, health, and marital status. Women, whether living alone or with others, experienced increases in loneliness; women living alone reported the greatest increase in loneliness. Men living alone reported high levels of loneliness prior to the pandemic, but only a slight increase over time. These analyses, which demonstrate that loneliness changed at the onset of the pandemic as a function of gender and living arrangement identify older people most likely to benefit from intervention.

Effects of peri-traumatic stress experienced during Hurricane Sandy on functional limitation trajectories for older men and women

BACKGROUND AND METHODS

Although the short-term effects of disasters on the physical health of mid-life and older people have been documented, little is understood about the long-term effects that disasters have on the physical health of these people. Based on the environmental docility hypothesis and research regarding gender effects on functional limitations and disaster, our analyses examined the effects of peri-traumatic stress experienced during Hurricane Sandy using longitudinal data from 5688 people aged 50 and older collected over six waves (2006-2019).

RESULTS

We found that functional limitations follow three trajectories, with people in each group having a significant linear increase over time and all but the highest functioning people also having a significant quadratic effect, indicating that the linear increase peaked post-Hurricane and then slowed in later waves.

CONCLUSION

Consistent with the environmental docility hypothesis, peri-traumatic stress had its greatest impact on people with more functional limitations before the hurricane. Men experiencing peri-traumatic stress during Hurricane Sandy were more likely to experience an increase in functional limitations than women. These findings, which identify people most likely to experience long-term health effects following a disaster, can be used to inform health policies before, during, and after disaster strikes.

Marital Status, Marital Transitions, and Sleep Quality in Mid to Late Life

This study aimed to understand how marital status and marital transitions were related to sleep quality in mid to late life and whether these findings differed by gender. Data from 2,872 participants 50-74 years old from the ORANJ BOWL^SM, a longitudinal panel study in New Jersey, were used. Marital status and sleep quality were examined in two waves that were approximately 10 years apart. Individuals in a significant romantic relationship and women had worse sleep quality than those in other marital status groups and men. Compared to individuals who remained married, individuals who remained divorced or widowed or who became widowed had better sleep quality, whereas those who became divorced had worse sleep quality; individuals who transitioned into marriage had better sleep quality than those who remained divorced or widowed. Marital status and gender appear important for at least some indices of sleep quality, an important predictor of late-life health.