Validation of Fuld object memory evaluation for the detection of dementia in nursing home residents

Depressive Symptoms as a Predictor of Memory Complaints in the PRISM Sample

Objectives

The current study investigated baseline and longitudinal relationships between memory complaints, depressive symptoms, and cognition in older adults.

Method

Using the sample from the Personal Reminder Information and Social Management trial, we generated path models predicting self-rated memory complaints measured by the Memory Functioning Questionnaire (MFQ).

Results

Our baseline models showed that more depressive symptoms were associated with reporting more frequent forgetting incidents and a greater decline in memory function. The baseline models also revealed that higher scores in a latent cognitive function were associated with reporting a greater decline in memory functioning and a greater use of mnemonics. However, cognitive predictors did not mediate the baseline associations between the MFQ measures and depressive symptoms. Further, these predictors were not able to directly predict the 12-month MFQ measures over and above the baseline effects. Including personality traits (neuroticism and conscientiousness) did not significantly affect the models.

Discussion

Our results suggest that memory complaints about frequency of forgetting can be the most reliable indicator of depression risk among the four factors in the MFQ. We discuss theoretical implications for longitudinal relationships between memory complaints, depressive symptoms, and cognitive function in older adults.

Psychometric validation of Fuld Object Memory Evaluation in older adults with cognitive impairments

OBJECTIVES

Mild cognitive impairment (MCI) refers to an early but abnormal state of cognitive impairment with minimal functional impairment. The present study aimed to evaluate the validity of Fuld Object Memory Evaluation (FOME) as a measure of episodic memory function.

METHOD

The study sample included 204 Chinese older adults with cognitive impairments. The participants completed five recall trials and a delayed trial in FOME, neurocognitive measures on digit spans and trail making, and daily functioning. Discriminative power of FOME to differentiate between MCI and dementia was inspected via receiver operating characteristic curve analysis.

RESULTS

FOME showed good test-retest reliability and convergent validity with digit spans and trail making. Controlling for gender, age, and education, lower levels of FOME total retrieval, verbal fluency, and daily functioning significantly predicted a higher likelihood of dementia compared to MCI. The optimal cut-off scores for total retrieval, verbal fluency, and daily functioning to differentiate dementia were 37/38 (77% sensitivity and 83% specificity), 28/29 (85% sensitivity and 72% specificity), and 14/15 (92% sensitivity and 78% specificity), respectively, in the younger subgroup. The corresponding figures were 34/35 (69% sensitivity and 76% specificity) and 27/28 (92% sensitivity and 62% specificity), and 11/12 (74% sensitivity and 80% specificity), respectively, in the older subgroup.

CONCLUSION

The findings support the FOME as a valid assessment tool of episodic memory function in older Chinese adults. The combined use of FOME and daily functioning is recommended to distinguish persons with dementia from MCI.

Normative data of Fuld Object Memory Evaluation test for brazilian elderly population

ABSTRACT Objective This study aims to present normative data for Fuld Object Memory Evaluation test stratified by sex, gender, age, and education for the Brazilian elderly population. Method We evaluated 2.132 healthy elderly both genders, with a mean age of 70.30 years (± 7.28) from two community-based samples in Brazil drawn from different economic areas who were screened with cognitive and functional tests and the memory test. Statistical analyses were performed by independent t-test, one-way analysis of variance and multiple linear regression. Results Statistical analyses showed that memory scores tend to improve significantly with increasing years of education and decrease significantly as age increased. Conclusion We conclude that gender, education and age had effect on the Fuld Object Memory Evaluation performance in this Brazilian community-based sample.

Octogenarian and centenarian performance on the Fuld Object Memory Evaluation

The Fuld Object Memory Evaluation (FOME) has considerable utility for cognitive assessment in older adults, but there are few normative data, particularly for the oldest old. In this study, 80 octogenarians and 244 centenarians from the Georgia Centenarian Study completed the FOME. Total and trial-to-trial performance on the storage, retrieval, repeated retrieval, and ineffective reminder indices were assessed. Additional data stratified by age group, education, and cognitive impairment are provided in the Supplemental data. Octogenarians performed significantly better than centenarians on all FOME measures. Neither age group benefitted from additional learning trials beyond Trial 3 for storage and Trial 2 for retention and retrieval. Ineffective reminders showed no change across learning trials for octogenarians, while centenarians improved only between Trials 1 and 2. This minimal improvement past Trial 2 indicates that older adults might benefit from a truncated version of the test that does not include trials three through five, with the added benefit of reducing testing burden in this population.

Review of brief cognitive tests for patients with suspected dementia

BACKGROUND

As the population ages, it is increasingly important to use effective short cognitive tests for suspected dementia. We aimed to review systematically brief cognitive tests for suspected dementia and report on their validation in different settings, to help clinicians choose rapid and appropriate tests.

METHODS

Electronic search for face-to-face sensitive and specific cognitive tests for people with suspected dementia, taking ≤ 20 minutes, providing quantitative psychometric data.

RESULTS

22 tests fitted criteria. Mini-Mental State Examination (MMSE) and Hopkins Verbal Learning Test (HVLT) had good psychometric properties in primary care. In the secondary care settings, MMSE has considerable data but lacks sensitivity. 6-Item Cognitive Impairment Test (6CIT), Brief Alzheimer's Screen, HVLT, and 7 Minute Screen have good properties for detecting dementia but need further validation. Addenbrooke's Cognitive Examination (ACE) and Montreal Cognitive Assessment are effective to detect dementia with Parkinson's disease and Addenbrooke's Cognitive Examination-Revised (ACE-R) is useful for all dementias when shorter tests are inconclusive. Rowland Universal Dementia Assessment scale (RUDAS) is useful when literacy is low. Tests such as Test for Early Detection of Dementia, Test Your Memory, Cognitive Assessment Screening Test (CAST) and the recently developed ACE-III show promise but need validation in different settings, populations, and dementia subtypes. Validation of tests such as 6CIT, Abbreviated Mental Test is also needed for dementia screening in acute hospital settings.

CONCLUSIONS

Practitioners should use tests as appropriate to the setting and individual patient. More validation of available tests is needed rather than development of new ones.

Bias in discriminating very mild dementia for older adults with different levels of education in Hong Kong

BACKGROUND

Education has a profound effect on older adults' cognitive performance. In Hong Kong, some dementia screening tasks were originally designed for developed population with, on average, higher education.

METHODS

We compared the screening power of these tasks for Chinese older adults with different levels of education. Community-dwelling older adults who were healthy (N = 383) and with very mild dementia (N = 405) performed the following tasks: Mini-Mental State Examination, Alzheimer's Disease Assessment Scale-Cognitive subscales, Verbal Fluency, Abstract Thinking, and Visual/Digit Span. Logistic regression was used to examine the power of these tasks to predict Clinical Dementia Rating (CDR 0.5 vs. 0).

RESULTS

Logistic regression analysis showed that while the screening power of the total scores in all tasks was similar for high and low education groups, there were education biases in some items of these tasks.

CONCLUSION

The differential screening power in high and low education groups was not identical across items in some tasks. Thus, in cognitive assessments, we should exercise great caution when using these potentially biased items for older adults with limited education.