The Montreal Cognitive Assessment (MoCA) is a rapid screening tool used to detect mild cognitive impairment (MCI), assessing cognitive domains (visuospatial skills, executive function, memory, attention, language, and orientation). Individual test scores are summed into a total score from 0 (worst) to 30 (best). Individuals may achieve any score within that range. Individual test scores are not reported.

Scores:

26-30 considered Normal; 19-25 may suggest Mild Cognitive Impairment (MCI); 10-18 can suggest moderate impairment; < 10 may indicate severe impairment

Cognitive status assessed by traditional neuropsychological tests measuring visuospatial skills, executive function, memory, attention, language, orientation. Results compared to age-adjusted norms, reviewed by expert neuropsychologists for consensus classification, per Univ of Pittsburgh Alzheimers Disease Research Ctr and Ntl Alzheimers Coordinating Ctr, and Alzheimer's Assoc criteria. Final classification adjusted per clinical judgment.

Impaired: ≥3 scores ≥2SD below norms Indeterminate: ≤2 scores >1SD below norms Normal: Neither criteria met. Total and domain scores not reported.

Cognitive status assessed by traditional neuropsychological tests measuring visuospatial skills, executive function, memory, attention, language, orientation. Results compared to age-adjusted norms, reviewed by expert neuropsychologists for consensus classification, per Univ of Pittsburgh Alzheimers Disease Research Ctr and Ntl Alzheimers Coordinating Ctr, and Alzheimer's Assoc criteria. Final classification adjusted per clinical judgment.

Impaired: ≥3 scores ≥2SD below norms Indeterminate: ≤2 scores >1SD below norms Normal: Neither criteria met. Total and domain scores not reported.

The Montreal Cognitive Assessment (MoCA) is a rapid screening tool to detect mild cognitive impairment (MCI), assessing cognitive domains (visuospatial skills, executive function, memory, attention, language, and orientation). Individual subtest scores are summed for a total score from 0 (worst) to 30 (best). Individuals may achieve any score within that range. Individual subtest scores are not reported.

Score range with Classification: 26-30 considered Normal; 19-25 may suggest Mild Cognitive Impairment (MCI); 10-18 can suggest moderate impairment; < 10 may indicate severe impairment

CAMCI (Computer Assessment of Memory and Cognitive Impairment) is a screening tool for assessment of cognitive status using neuropsychological tests of memory, attention, executive ability modified for computer administration, and a Virtual Environment task assessing cognitive areas not easily assessed by paper tests. Score is a weighted sum of individual task scores converted to total score from 0-worst to 50-best. Intended use provides single score/cognitive classification. Individual task scores not reported. Score Range with Classification: 21-50=Normal, 16-20=Indeterminate, 0-15=Impaired

The Wide Range Achievement Test, 5th Edition (WRAT5), Word Reading test assess fundamental reading skills and can indicate cognitive impairments affecting academic performance. Participant reads aloud a list of letters (15) and words (55). One point for each letter or word read correctly. Higher scores are better. Score range 0-70 points.

The Digit Span Forward (DSF) (Wechsler et al., 1987), part of the Wechsler Adult Intelligence Scale, 4th Edition (WAIS-IV) evaluates short-term memory and attention. Examiner reads a span of digits at a rate of one digit per second. Participant must repeat the digits back to the examiner in sequential (forward) order. Two trials are presented at each span length. The length of span is increased (Max Span 0-9) if at least 1 correct response given at the current span length. A point is awarded for each correct recall. Score is the total points obtained (0-16). Higher scores are better.

The Digit Span Backward (DSB) (Wechsler et al., 1987), part of the Wechsler Adult Intelligence Scale, 4th Edition (WAIS-IV) evaluates working memory and cognitive flexibility. Examiner reads a span of digits at a rate of one digit per second. Participant must repeat the digits back to the examiner in reverse (backward) order. Two trials are presented at each span length. The length of span is increased (Max Span 0-8) if at least 1 correct response given at the current span length. A point is awarded for each correct recall. Score is the total points obtained (0-16). Higher scores are better.

The Logical Memory I, part of the Wechsler Memory Scale, 4th Edition (WMS-IV), Wechsler (1987), evaluates immediate verbal memory recall and the ability to encode and retrieve information quickly. Participant is read 2 short stories selected by age (Adult=60-69 years old, Older Adult=65-90 years old) and asked to recall as many details as possible immediately after hearing each story. Score is the number of details correctly recalled (1 point per correct detail.) Scores range based on story complexity (0-14, or 0-25). Story scores are summed for a Total Score.

The standardized test is provided in two forms that are selected based on participant age and ability. The age groups for this study are adults ages 60 to 69 years old and older adults ages 65 to 90 years old. There were 355 participants assigned to the adult group and 409 participants assigned to the older adult group.

The Logical Memory II, part of the Wechsler Memory Scale, 4th Edition (WMS-IV), Wechsler (1987), evaluates delayed memory recall, which involves the ability to store information and retrieve it after a period of time. Participant is read 2 short stories selected by age (Adult=60-69 years old, Older Adult=65-90 years old). After a delay of ~30 minutes, participant is asked to recall as many details as possible. Score is the number of details correctly recalled (1 point per correct detail.) Scores range based on story complexity (0-14, or 0-25). Story scores are summed for a Total Score.

The standardized test is provided in two forms that are selected based on participant age and ability. The age groups for this study are adults ages 60 to 69 years old and older adults ages 65 to 90 years old. There were 355 participants assigned to the adult group and 409 participants assigned to the older adult group.

The Rey-Osterrieth Complex Figure Test (ROCF) (Meyers and Meyers, 1995), evaluates visuospatial abilities, memory, attention, and executive functions. Participant is presented with a complex geometric figure to study then reproduce the figure on a blank sheet of paper. After ~3 minutes (immediate recall) and ~30 minutes (delayed recall) participant must reproduce the figure again from memory. Each reproduction is scored based on 18 design elements with 0, .5, 1, or 2 points awarded per element based on element accuracy and placement.

The Letter Fluency (F-A-S) Test, a subtest of the Neurosensory Center Comprehensive Examination for Aphasia (Spreen and Benton, 1977) evaluates phonemic verbal fluency and can assess language abilities, executive functions, and cognitive flexibility. Participant is asked to generate as many words as possible that begin with the letters "F," "A," and "S" within 60 seconds for each letter. Score is total words generated (1 point per unique word) within the time limit. Higher scores are better. There is no fixed upper limit of the score.

The Semantic Fluency Test (animals) (Spreen and Strauss, 1997), evaluates semantic memory, language abilities, and executive functions, including cognitive flexibility and retrieval from long-term memory. Participant is asked to name as many animals as they can within 60 seconds. Score is total animal names generated (1 point per unique animal named) within the time limit. Higher scores are better. There is no fixed upper limit of the score.

The Hopkins Verbal Learning Test - Revised (HVLT-R) evaluates verbal learning and memory. Participant is presented with a list of 12 nouns, 4 words each from 3 semantic categories and asked to recall them over a series of 3 Learning Trials. Score ranges 0-12 on each trial (1 point per each correct word). The score from each trial is summed for a Learning Trials Total Recall Score (0-36). After a 20 minute delay, the participant performs a final recall for a Delayed Recall Score (0-12). Higher is better for all scores. Retention % = Delayed Recall score / high score from Learning Trials 2 or 3.

The Hopkins Verbal Learning Test - Revised (HVLT-R) evaluates verbal learning and memory. Participant is presented 12 nouns, 4 words each from 3 semantic categories and asked to recall them over a series of 3 Learning Trials. Score ranges 0-12 on each trial (1 point per correct word). The score from each trial is summed for a Learning Trials Total Recall Score (0-36). After a 20 minute delay, participant performs a final recall for a Delayed Recall Score (0-12). Retention % = Delayed Recall score / high score from Learning Trials 2 or 3. Higher Retention % is better, as is not capped to 100%.

The Trail Making Test (TMT) (Reitan et al., 1958) evaluates processing speed, attention, and executive function. Participant draws lines as instructed to connect 25 circles labeled with numbers or letters without lifting the pen. One point awarded per correct line drawn (0-24, higher is better.) Task completion time is recorded in seconds. In Part A participant has 150 seconds to connect the numbered circles in ascending order (1-2-3, etc.). In Part B participant has 300 seconds to connect 25 circles alternating between numbers and letters in ascending order (1-A-2-B, etc.)

The Hopkins Verbal Learning Test - Revised (HVLT-R) evaluates verbal learning and memory. Participant is presented 12 nouns and asked to recall them over 3 Learning Trials followed by a 20 minute Delayed Recall. Participant is then presented 12 target + 12 distractor words (6 semantically related, 6 semantically unrelated) and asked to recognize any prior words presented. Retention % = Delayed Recall score divided by high score from Learning Trials 2 or 3. Retention Discrimination Index (-12 to 12) is calculated as number of true-positives - number of false positives. Higher scores are better.

The Trail Making Test (TMT) (Reitan et al., 1958) evaluates processing speed, attention, and executive function. Participant draws lines as instructed to connect 25 circles labeled with numbers or letters without lifting the pen. One point awarded per correct line drawn (0-24, higher is better.) Task completion time is recorded in seconds. In Part A participant has 150 seconds to connect the numbered circles in ascending order (1-2-3, etc.). In Part B participant has 300 seconds to connect 25 circles alternating between numbers and letters in ascending order (1-A-2-B, etc.)

The Digit Symbol Substitution Test (DSST) part of the Wechsler Adult Intelligence Scale-IV (WAIS-IV) (Wechsler et al., 1987) evaluates processing speed, attention, and motor skills. Performance correlates with real-world functional outcomes e.g., ability to accomplish everyday tasks. Participant is given a sheet of numbers and a key showing a unique symbol corresponding to each number. Participants match symbols to numbers by copying the correct symbol into spaces below each number within 120 seconds. Higher scores are better. Score range 0-135, 1 point for each correct symbol matched.

The Boston Naming Test (BNT) 2nd edition (Kaplan et al., 2000) evaluates language abilities, specifically confrontation naming. Participant is shown 60 black-and-white line drawings of objects, one at a time for up to 20 seconds and asked to name each object. If participant struggles to name an object, examiner can provide a stimulus cue (descriptive) and a phonemic cue (sound-based). Score is the number of correct responses, including those given after a stimulus cue but excluding those given after a phonemic cue. Scores range from 0-60. Higher scores are better.

Participants classified as Indeterminate through clinical adjudication.

Participants classified as Impaired through clinical adjudication.

Linear regression agreement analysis (linear regression equation and confidence intervals, and Pearson correlation) between CAMCI and Montreal Cognitive Assessment (MoCA) for the Test group. Comparison (positive and negative percent agreement, confidence interval estimates, and quadratic weighted kappa) between CAMCI and Montreal Cognitive Assessment (MoCA) for the Test group.

CAMCI Scores ranged from 0-50, higher scores mean better outcome. MoCA scores ranged from 0-30, higher scores mean better outcome.

Power calculations for agreement were assessed for Cohen Kappa's assuming at least 75% Normals with 80% power and .05 significance level, finding that we were sufficiently powered with at least 120 individuals in the test set, for a Kappa of at least 0.3.

Test group. Of 773 participants enrolled, 9 were excluded, leaving 764 for analysis. Participants who completed CAMCI version A at baseline (N=680) were assigned through stratified random sampling to Training (75%) or Test (25%) groups, balanced by age, education, race, ethnicity, gender, and cognitive classification. The groups were further adjusted to match the racial and ethnic distribution in the U.S., resulting in a Training group (N=374) and a Test group (N=126).

To derive a measure of reliable change that addressed unbalanced repeated measures and unequal timing of repeat measures, we built a two-stage hierarchical model for longitudinal data using Bayesian methods. Additionally, the standard deviation (SD) was allowed to change over time, which allowed for computation of an intraclass correlation coefficient (ICC), a measure of reliability. With an SD and ICC that can change over time, reliable change was computed for follow-up months from 3 to 24, in groups of 3. Reliable change was then adjusted for practice effects according to Duff, K. (2012).

Computer Assessment of Memory and Cognitive Impairment (CAMCI) scores were calculated as the weighted sum of individual task scores converted to a total score from 0-worst to 50-best. As the intended use of CAMCI is to provide a single score of overall cognitive status, individual task scores were not reported.

Individuals who received a consistent version of CAMCI across all visits (i.e., CAMCI-A), and were classified as Normal through clinical adjudication at all visits.