Monthly At-Home Computerized Cognitive Testing to Detect Diminished Practice Effects in Preclinical Alzheimer's Disease

Introduction: We investigated whether monthly assessments of a computerized cognitive composite (C3) could aid in the detection of differences in practice effects (PE) in clinically unimpaired (CU) older adults, and whether diminished PE were associated with Alzheimer's disease (AD) biomarkers and annual cognitive decline.Materials and Methods:N = 114 CU participants (age 77.6 ± 5.0, 61% female, MMSE 29 ± 1.2) from the Harvard Aging Brain Study completed the self-administered C3 monthly, at-home, on an iPad for one year. At baseline, participants underwent in-clinic Preclinical Alzheimer's Cognitive Composite-5 (PACC5) testing, and a subsample (n = 72, age = 77.8 ± 4.9, 59% female, MMSE 29 ± 1.3) had 1-year follow-up in-clinic PACC5 testing available. Participants had undergone PIB-PET imaging (0.99 ± 1.6 years before at-home baseline) and Flortaucipir PET imaging (n = 105, 0.62 ± 1.1 years before at-home baseline). Linear mixed models were used to investigate change over months on the C3 adjusting for age, sex, and years of education, and to extract individual covariate-adjusted slopes over the first 3 months. We investigated the association of 3-month C3 slopes with global amyloid burden and tau deposition in eight predefined regions of interest, and conducted Receiver Operating Characteristic analyses to examine how accurately 3-month C3 slopes could identify individuals that showed >0.10 SD annual decline on the PACC-5.Results: Overall, individuals improved on all C3 measures over 12 months (β = 0.23, 95% CI [0.21–0.25], p < 0.001), but improvement over the first 3 months was greatest (β = 0.68, 95% CI [0.59–0.77], p < 0.001), suggesting stronger PE over initial repeated exposures. However, lower PE over 3 months were associated with more global amyloid burden (r = −0.20, 95% CI [−0.38 – −0.01], p = 0.049) and tau deposition in the entorhinal cortex (r = −0.38, 95% CI [−0.54 – −0.19], p < 0.001) and inferior-temporal lobe (r = −0.23, 95% CI [−0.41 – −0.02], p = 0.03). 3-month C3 slopes exhibited good discriminative ability to identify PACC-5 decliners (AUC 0.91, 95% CI [0.84–0.98]), which was better than baseline C3 (p < 0.001) and baseline PACC-5 scores (p = 0.02).Conclusion: While PE are commonly observed among CU adults, diminished PE over monthly cognitive testing are associated with greater AD biomarker burden and cognitive decline. Our findings imply that unsupervised computerized testing using monthly retest paradigms can provide rapid detection of diminished PE indicative of future cognitive decline in preclinical AD.

Using Telehealth-Delivered Procedures to Collect a Parent-Implemented Expressive Language Sampling Narrative Task in Monolingual and Bilingual Families With Autism Spectrum Disorder: A Pilot Study

Language impairments are frequent, severe, and of prognostic value in autism spectrum disorder (ASD). Unfortunately, the evaluation of the efficacy of treatments targeting the language skills of those with ASD continues to be hindered by a lack of psychometrically sound outcome measures. Expressive Language Sampling (ELS) procedures offer a promising alternative to norm-referenced standardized tests for assessing expressive language in treatment studies. Until now, however, research on the validity and utility of ELS as outcome measures has been limited to administrations by a trained professional in a clinic setting and to use with English-speaking families. These limitations are a barrier for many families accessing the benefits of participation in treatment studies. The current study examines the feasibility of teaching native English-speaking parents (NESP) and native Spanish-speaking parents (NSSP) how to administer the ELS narrative task (ELS-N) to their sons and daughters with ASD (between ages 6 and 21) at home through telehealth-delivered procedures. The parent training was provided in the primary language of the participating parent (i.e., 11 NSSP and 11 NESP) and administered by the parent to the youth in the language that the parent reported to use to communicate with the youth at home (i.e., 9 Spanish and 13 English). Families were able to choose between using their own technology or be provided with the technology needed for participation. Of the 19 parents who completed the training, 16 learned to administer the ELS-N procedures. In addition, strong test-retest reliability and no practice effects over the 4-week interval were observed for ELS-N derived youth outcome measures (i.e., talkativeness, vocabulary, syntax, dysfluency, and intelligibility) for both NSSP and NESP. Results from this pilot study suggest that the home-based parent-implemented ELS-N procedures can be learned and administered at acceptable levels of fidelity by parents, with good test-retest reliability and limited practice effects observed in terms of outcome measures for youth with ASD. Implications for treatment studies and future directions are discussed.

Effects of Repeated Administration and Comparability of Alternate Forms for the Global Neuropsychological Assessment (GNA)

The Global Neuropsychological Assessment (GNA) is an extremely brief battery of cognitive tasks assessing episodic memory, processing speed, working memory, verbal fluency, executive function, and mood. It can be given in under 15 minutes, has five alternate forms, and does not require an examinee to be literate. The purpose of this study was to quantify practice effects over repeated administrations and assess comparability of the GNA’s five alternate forms, preparing the battery for repeated administration in research and clinical settings. Forty participants each completed all five GNA forms at weekly intervals following a Latin square design (i.e., each form was administered at every position in the sequence an equal number of times). In a cognitively intact population, practice effects of 0.56 to 1.06 SD were observed across GNA measures when comparing the first and fifth administration. Most GNA tests showed nonsignificant interform differences with cross-form means differing by 0.35 SD or less, with the exception of modest but statistically significant interform differences for the GNA Story Memory subtest across all five forms. However, post hoc analysis identified clusters of two and three Story Memory alternate forms that were equivalent.

Changes in Processing Speed, Cognitive Flexibility, and Selective Attention over a Four-Week Treatment Period in Inpatients with Moderate to Severe Depression

Cognitive dysfunction among individuals with depression is associated with reduced functional status, and cognitive improvement is often an important treatment goal. We compared changes in cognitive performance over four weeks among 45 inpatients with a unipolar depressive disorder completing inpatient treatment to that of 20 controls on measures of processing speed and set-shifting (Trail Making Test), as well as selective attention (Test d2). In the patients, depressive symptoms improved significantly and with a large effect over the treatment period ( d = 1.22–1.81). Among the three cognitive domains examined, the most pronounced reductions among patients compared to controls were observed in cognitive flexibility (Group effect: ηp2 = .04). The effect of Group; however, was not significant. Likewise, there were no significant improvements in cognitive flexibility over time, and changes in cognitive flexibility over the four-week period did not differ between the two groups (Group x Time interaction). Performances in selective attention and processing speed improved over the four-week period, though neither the effect of Group nor the interaction (Group x Time) was significant regarding these performances. Change in cognitive performance was not associated with changes in symptom severity (incl. remission status). Taken together, the significant improvements in selective attention and processing speed were largely attributable to practice effects. Our findings lend further support to the notion that cognitive flexibility, selective attention, and processing speed are independent of improvements in depressive symptoms. This study underscores the importance of including comparison groups to control for practice effects when examining cognitive change, and providing treatments specifically aimed at improving cognitive symptoms.