DISSOCIATING RETEST EFFECTS FROM DEVELOPMENTAL CHANGE FOR PREDICTING COGNITIVE STATUS

Abstract In longitudinal designs, unadjusted retest effects can confound developmental change estimates. This study utilized a measurement burst design and three-level multilevel modeling to a) independently parameterize short-term retest and long-term developmental change and b) employ these estimates as predictors of cognitive status at long-term follow-ups. Using data from Project MIND , participants (N=304; aged 64-92 years) were assessed across biweekly sessions nested within annual bursts (spanning up to 17 total assessments over four years). Cognitive impairment no dementia (CIND) status was classified at Years 4 (the final burst assessment) and 8 (the study end date). Response time inconsistencies (RTI) were computed to index intraindividual variability across RT trials of a one-back response time (BRT) task. Three-level multilevel models simultaneously yet independently estimated BRT RTI change across weeks and years, indexing short-term retest and long-term developmental change, respectively. Individual slope estimates were extracted and utilized in multinomial logistic regression models contrasting short- vs. long-term RTI change as predictors of long-term cognitive status. Results from the three-level models indicated that retest and developmental slopes yielded non-redundant sources of variance, providing unique estimates of change that would otherwise be confounded. Further, short- and long-term RTI differentially predicted cognitive status at Years 4 and 8; failing to benefit from retest effects on the BRT task was associated with increased likelihood of cognitive impairment. This innovative approach to parameterizing retest effects can reduce systematic bias in estimates of long-term developmental change, as well as highlight the utility of retest effects as predictors of cognitive health.

and cognitive functioning, but that these relationships are not robust across time.Despite support provided by unpaid family caregivers, African American older adults with vision impairment are at an increased risk for concurrent social and cognitive challenges.It may be that family support of those with sensory impairments helps so that these impairments aren't related to social isolation or cognitive functioning across time.Researchers and clinicians could benefit older African Americans with sensory impairments by providing and encouraging support during early stages of vision loss.

CHILDHOOD MALTREATMENT AND PROSOCIAL BEHAVIOR: A QUALITATIVE COMPARATIVE STUDY OF IRISH OLDER ADULT SURVIVORS
Shauna Rohner 1 , Aileen Salas 2 , Alan Carr 3 , and Myriam Thoma 4 , 1. University of Zürich,Zürich,Zurich,Switzerland,2. University of Zürich,Zurich,Zurich,Switzerland,3. University College Dublin,Dublin,Dublin,Ireland,4. University of Zurich,Zurich,Zurich,Switzerland Although childhood maltreatment can have lasting effects into later life, positive outcomes have also been observed, including an increased tendency towards prosocial behavior.However, little is known about the link between childhood maltreatment and later life prosocial behavior.Therefore, this study aimed to explore older adult's experiences of childhood maltreatment and identify mechanisms linked to prosocial behavior in later life.The individual level, but also broader cultural and contextual mechanisms, were considered by comparing two adversity contexts and applying conceptual frameworks (socio-interpersonal framework model of trauma and recovery, motivational process model of altruism born of suffering).Semi-structured interviews (60-120 minutes) were conducted with 29 Irish (older) adult survivors of childhood maltreatment: 17 institutional (welfare care) abuse survivors (mean age: 61 years, range: 50-77), 12 familial abuse survivors (mean age: 58 years, range: 51-72).Interviews were analyzed using Framework Analysis.In both groups at the individual level, enhanced empathy, amelioration, and identity-related mechanisms were linked to prosocial behavior, with connections to caring roles and coping strategies from childhood.On a social contexts level, the limited resources or opportunities for help in childhood, and the social norms and beliefs of that time, influenced participants' motivation to help others in later life.Group-specific mechanisms were also observed, such as compassion fatigue in the familial sample; and denouncing detrimental societal values in the institutional sample.The identification of individual, adversity-context, and culturespecific mechanisms linked to later-life prosocial behavior can promote a greater understanding of resilience and adaptability in older adult survivors of childhood maltreatment.The enrollment of human subjects is crucial for the success of clinical trials.In the ongoing "Reducing the Risk of Dementia through Deprescribing" trial, the initial approach for enrolling subjects did not meet expected goals in the first 6 months, creating the need for innovative nudge strategies.We used an Agile methodology as the framework to understand the problem, then find and implement a solution.

USING AGILE METHODOLOGY AND NUDGE STRATEGIES TO IMPROVE ENROLLMENT IN CLINICAL TRIALS
Our study aimed to examine the effectiveness of utilizing a texting nudge to enhance post-agreement recruitment of subjects with cognitive impairments.Prior to enrollment, eligible potential participants were contacted using a texting nudge.Potential participants received a second contact call to remind subjects of the enrollment appointment, introduce the person and the phone number that would call them, and the option of confirming or rescheduling.During the 1-week text-message experiment, 8 out of 9 subjects who agreed to participate in the study and received the text message enrolled, yielding an 89% post-agreement enrollment rate compared to a baseline rate of 44% prior to introducing this nudge.After implementing into the standard operating procedures, the 6-month average rate of enrollment among those that agreed rose to 80%, nearly doubling the rate from the first 6 months of the study and quadrupling the number enrolled each month.Inadequate recruitment has necessitated the use of innovative recruitment methods.Using the Agile problem-solving mindset, the texting nudge was developed to leverage the behavioral influences of the messenger, social commitments, priming and affect to increase subject enrollment.In longitudinal designs, unadjusted retest effects can confound developmental change estimates.This study utilized a measurement burst design and three-level multilevel modeling to a) independently parameterize short-term retest and long-term developmental change and b) employ these estimates as predictors of cognitive status at long-term follow-ups.Using data from Project MIND , participants (N=304; aged 64-92 years) were assessed across biweekly sessions nested within annual bursts (spanning up to 17 total assessments over four years).Cognitive impairment no dementia (CIND) status was classified at Years 4 (the final burst assessment) and 8 (the study end date).Response time inconsistencies (RTI) were computed to index intraindividual variability across RT trials of a one-back response time (BRT) task.Three-level multilevel models simultaneously yet independently estimated BRT RTI change across weeks and years, indexing short-term retest and long-term developmental change, respectively.Individual slope estimates were extracted and utilized in multinomial logistic regression models contrasting short-vs.long-term RTI change as predictors of long-term cognitive status.Results from the three-level models indicated that retest and developmental slopes yielded non-redundant sources of variance, providing unique estimates of change that would otherwise be confounded.Further, short-and long-term RTI differentially predicted cognitive status at Years 4 and 8; failing to benefit from retest effects on the BRT task was associated with increased likelihood of cognitive impairment.This innovative approach to parameterizing retest effects can reduce systematic bias in estimates of long-term developmental change, as well as highlight the utility of retest effects as predictors of cognitive health.Over the past years, recruitment of participants for behavioral and biomedical research through the internet has become more popular.Online has become an advantageous approach to recruitment, especially since the covid pandemic has posed a great challenge to most in-person research activities.Typically, internet-based recruitment strategies include website posting, emailing list of potential participants as well as using social media such as Facebook, Twitter, and Instagram to deliver recruitment information to groups who are often unrepresented in research.Although this mechanism has reduced barriers to participation, it has posed serious threats to data quality and validity.For example, some studies on data validity estimate that up to 90% of online survey responses are fraudulent when they rely on screening questions and CAPTCHA alone.Others have shown that vetted panel data such as Mechanical Turk (mTurk), has high rates of participant misrepresentation.Therefore, the aim of this paper is to highlight the challenges associated with internet-based recruitment of family caregivers and describe strategies for researchers to ensure data integrity.We discuss multi-faceted approaches to detect and prevent fraudulent and suspicious activities such as duplicate and automated enrollment by software applications known as bots as well by fraudulent human participants.We discuss data on several strategies that have proven effective in our previous and ongoing trails.We will also demonstrate the need to implement several strategies and a "fail-safe" to detect fraud after enrollment.It is imperative that researchers understand the need to address these challenges to preserve data integrity and replicability.
h increased likelihood of cognitive impairment.This innovative approach to parameterizing retest effects can reduce systematic bias in estimates of long-term developmental change, as well as highlight the utility of retest effects as predictors of cognitive health.Over the past years, recruitment of participants for behavioral and biomedical research through the internet has become more popular.Online has become an advantageous approach to recruitment, especially since the covid pandemic has posed a great challenge to most in-person research activities.Typically, internet-based recruitment strategies include website posting, emailing list of potential participants as well as using social media such as Facebook, Twitter, and Instagram to deliver recruitment information to groups who are often unrepresented in research.Although this mechanism has reduced barriers to participati n, it has posed serious threats to data quality and validity.For example, some studies on data validity estimate that up to 90% of online survey responses are fraudulent when they rely on screening questions and CAPTCHA alone.Others have shown that vetted panel data such as Mechanical Turk (mTurk), has high rates of participant misrepresentation.Therefore, the aim of this paper is to highlight the challenges associated with internet-based recruitment of family caregivers and describe strategies for researchers to ensure data integrity.We discuss multi-faceted approaches to detect and prevent fraudulent and suspicious activities such as duplicate and automated enrollment by software applications known as bots as well by fraudulent human participants.We discuss data on several strategies that have proven effective in our previous and ongoing trails.We will also demonstrate the need to implement several strategies and a "fail-safe" to detect fraud after enrollment.It is imperative that researchers understand the need to address these challenges to preserve data integrity and replicability.


DISSOCIATING RETEST EFFECTS FROM DEVELOPMENTAL CHANGE FOR PREDICTING COGNITIVE STATUS


STRATEGIES TO ENSURE AUTHENTIC PARTICIPANTS AND VALID DATA WITH ONLINE RECRUITMENT OF FAMILY CAREGIVERS


HAIR CORTISOL FEASIBILITY AND DEMOGRAPHIC CORRELATES IN A SAMPLE OF OLDER ADULTS FROM PUERTO RICO

Erin Ballard 1 , Bulent Turan 2 , and Michael Crowe 3 , 1. University of Alabama at Birmingham, Hoover, Alabama, United States,2. Koc University,Sariyer,Istanbul,Turkey,Birmingham,Alabama,United States Hair cortisol is increasingly being used as a biomarker of chronic HPA-axis activation.Studies using older adults often exclude a substantial portion of participants due to insufficient hair or non-detectable cortisol levels, but do not provide details on correlates of these factors.We examined feasibility of hair measurement and cortisol detectability in an ongoing study of older adults in Puerto Rico.Among the first 537 participants in the current follow-up of the Puerto Rican Elder Health Conditions (PREHCO) study (now age 78-106 years old), approximately 11% of participants refused to give a hair sample and 20% did not have enough hair to sample.Women (13%) were significantly more likely than men (4%) to refuse hair collection.However, men (47.7%) were significantly more likely than women (4.8%) to not have enough hair.Of participants with enough hair to take a sample (n=372), 23% had non-detectable levels of cortisol.Black participants were the most likely to have non-detectable hair cortisol (43%), followed by multiracial mestizo (32%).The two most common racial categories in our sample, multiracial trigueño (23%), and white (17%) were the least likely to have non-detectable cortisol.In terms of hair products (including frequency of hair washing and use of conditioner, dye, or perm), only the use of chemical hair straighteners was associated with higher likelihood of non-detectable cortisol (38%).Findings underscore the importance of measuring hair products when examining hair cortisol in older adults and suggest that older black participants may be disproportionately excluded due to non-detectable cortisol levels.


TIME IS OF THE ESSENCE: RELIABLE MEASUREMENT OF HEART RATE VARIABILITY IN OLDER ADULTS

Cristina Pinheiro, Suzanne Segerstrom, and Justin Karr,


University of Kentucky, Lexington, Kentucky, United States

Heart rate variability (HRV) decreases with age and is an important correlate of psychosocial and physical health.Recommendations for the minimum duration of EKG to accurately derive HRV vary from 1 minute to several minutes.However, the definition of "accuracy" or reliability depends on study design, including whether the focus of the study is on stable or momentary between-person differences or within-person changes in HRV.In a sample of 216 older adults (Mage = 72.7,62.5% women), ECG was measured at 1000 Hz for 10 minutes every 6 months for up to 6.5 years.HRV was high-frequency power (0.15-0.40).A generalizability study determined the variance due to minute, occasion, person, and their interactions.The most variance was due to idiosyncratic occasion differences (46%), followed by stable person variance (18%).A decision study determined how accuracy in HRV measurement could be achieved.Between-person differences in HRV at a specific occasion could be reliably measured with 6 minutes of ECG (.80); within-person changes in HRV between occasions could be reliably measured with 3 minutes of ECG (.83).For stable individual differences, 10 or more occasions of 1-minute duration produced a more reliable estimate (.73) than increasing the length of a single ECG recording to 10 minutes (.25).The necessary



Peggy Bylund 1 , Jade Me

HAIR CORTISOL FEASIBILITY AND DEMOGRAPHIC CORRELATES IN A SAMPLE OF OLDER ADULTS FROM PUERTO RICO
Erin Ballard 1 , Bulent Turan 2 , and Michael Crowe 3 , 1. University of Alabama at Birmingham, Hoover, Alabama, United States,2. Koc University,Sariyer,Istanbul,Turkey,Birmingham,Alabama,United States Hair cortisol is increasingly being used as a biomarker of chronic HPA-axis activation.Studies using older adults often exclude a substantial portion of participants due to insufficient hair or non-detectable cortisol levels, but do not provide details on correlates of these factors.We examined feasibility of hair measurement and cortisol detectability in an ongoing study of older adults in Puerto Rico.Among the first 537 participants in the current follow-up of the Puerto Rican Elder Health Conditions (PREHCO) study (now age 78-106 years old), approximately 11% of participants refused to give a hair sample and 20% did not have enough hair to sample.Women (13%) were significantly more likely than men (4%) to refuse hair collection.However, men (47.7%) were significantly more likely than women (4.8%) to not have enough hair.Of participants with enough hair to take a sample (n=372), 23% had non-detectable levels of cortisol.Black participants were the most likely to have non-detectable hair cortisol (43%), followed by multiracial mestizo (32%).The two most common racial categories in our sample, multiracial trigueño (23%), and white (17%) were the least likely to have non-detectable cortisol.In terms of hair products (including frequency of hair washing and use of conditioner, dye, or perm), only the use of chemical hair straighteners was associated with higher likelihood of non-detectable cortisol (38%).Findings underscore the importance of measuring hair products when examining hair cortisol in older adults and suggest that older black participants may be disproportionately excluded due to non-detectable cortisol levels.

TIME IS OF THE ESSENCE: RELIABLE MEASUREMENT OF HEART RATE VARIABILITY IN OLDER ADULTS
Cristina Pinheiro, Suzanne Segerstrom, and Justin Karr,

University of Kentucky, Lexington, Kentucky, United States
Heart rate variability (HRV) decreases with age and is an important correlate of psychosocial and physical health.Recommendations for the minimum duration of EKG to accurately derive HRV vary from 1 minute to several minutes.However, the definition of "accuracy" or reliability depends on study design, including whether the focus of the study is on stable or momentary between-person differences or within-person changes in HRV.In a sample of 216 older adults (Mage = 72.7,62.5% women), ECG was measured at 1000 Hz for 10 minutes every 6 months for up to 6.5 years.HRV was high-frequency power (0.15-0.40).A generalizability study determined the variance due to minute, occasion, person, and their interactions.The most variance was due to idiosyncratic occasion differences (46%), followed by stable person variance (18%).A decision study determined how accuracy in HRV measurement could be achieved.Between-person differences in HRV at a specific occasion could be reliably measured with 6 minutes of ECG (.80); within-person changes in HRV between occasions could be reliably measured with 3 minutes of ECG (.83).For stable individual differences, 10 or more occasions of 1-minute duration produced a more reliable estimate (.73) than increasing the length of a single ECG recording to 10 minutes (.25).The necessary Peggy Bylund 1 , Jade Mehta 2 , Nandini Mathavan 3 , Kimberly Trowbridge 1 , Britain Taylor 4 , Husam El Sharu 5 , and Noll Campbell 6 , 1. Regenstrief Institute, Inc., Indianapolis, Indiana, United States, 2. Indiana University School of Medicine, Indianapolis, Indiana, United States, 3. Indiana University School of Medicine/Center for Health Innovation and Implementation Science, Indianapolis, Indiana, United States, 4. Indiana University Bloomington, Bloomington, Indiana, United States, 5. Indiana University, Indianapoilis, Indiana, United States, 6. Purdue University, Indianapolis, Indiana, United States

iversity School of Medicine, Indianapolis, Indiana, United States, 3. Indiana University School of
Medicine/Center for Health Innovation and Implementation Science, Indianapolis, Indiana, United States, 4. Indian