Cognitive Assessments and Questionnaires
There are a wide assortment of tests and questionnaires administered to ADNI participants. Before diving into the tests themselves we’ll lay out some context for how the tests are administered, and how the set of assessments has varied over time.
Order of test administration
This is laid out in the procedures manual for each phase. For convenience, the relevant excerpts are included here:
Changes between phases
The exact composition of the testing battery has varied a fair bit over time. A table summarizing the administration of different instruments is included here in table 1:
Table 1
Instruments administered to participants
| Instrument | ADNI1 | ADNIGO | ADNI2 | ADNI3 | ADNI4 |
|---|---|---|---|---|---|
| Abbreviated Multidimensional Acculturation Scale (AMAS; for Latino participants only) | x | ||||
| Alzheimer’s Disease Assessment Scale-Cog13 (w/ delayed recall and number cancellation; ADAS-Cog) | x | x | x | x | x |
| American National Adult Reading Test (AMNART) | x | x | |||
| Boston Naming Test | x | x | x | ||
| Brief Perceived Ethnic Discrimination Questionnaire - Community Version (PEDQ-CV) | x | ||||
| Category Fluency (Animals) | x | x | x | x | x |
| Category Fluency (Vegetables) | x | ||||
| Clinical Dementia Rating scale (CDR) | x | x | x | x | x |
| Clock Drawing | x | x | x | x | x |
| Cognitive Change Index (CCI) | x | x | |||
| Cogstate Brief Battery (CBB) | x | x | |||
| Digit span (forward/backward; collected under MOCA after ADNI1) | x | ||||
| Digit symbol substitution (collected under MOCA after ADNI1) | x | ||||
| Everyday Cognition 12-item (ECog 12-item) | x | ||||
| Everyday Cognition 39-item (ECog) | x | x | x | ||
| Financial Capacity Instrument short form (FCI) | x | ||||
| Geriatric Depression Scale (GDS) | x | x | x | x | x |
| Hollingshead Index | x | ||||
| Impact of Events (for amyloid PET disclosure only) | x | ||||
| Logical Memory I and II | x | x | x | x | x |
| Mini Mental State Examination (MMSE) | x | x | x | x | x |
| Modified Hachinski Ischemic Score | x | x | x | x | |
| Montreal Cognitive Assessment (MoCA) | x | x | x | x | |
| Multi-Lingual Naming Test (MINT) | x | x | |||
| Perceived Stress Scale (PSS) | x | ||||
| Rey Auditory Verbal Learning Test (AVLT) | x | x | x | x | x |
| Ryff Purpose of Life (for amyloid PET disclosure only) | x | ||||
| Trails A & B | x | x | x | x | x |
| Word Accentuation Test - Chicago (WAT; for Spanish-language participants in place of AMNART) | x |
Note that there are other questionnaires related to non-biological measures, such as the ADI, RUCC, RUCA and others, that are covered in more detail in the ‘subject characteristics’ chapter of this document. There is also a set of assessments that are administered remotely as part of the ADNI4 Remote Digital cohort, which are covered in the corresponding chapter.
Table 1, above, covers instruments that are administered specifically to the participant. ADNI participants are required to have a study partner. The following table gives the instruments that either have an informant report version (the results of which are available) or rely on informant report in some way for the scoring procedure
Table 2:
instruments administered to study partners
| Instrument | ADNI1 | ADNIGO | ADNI2 | ADNI3 | ADNI4 |
|---|---|---|---|---|---|
| Functional Activities Questionnaire (FAQ) | x | x | x | x | x |
| Clinical Dementia Rating (CDR) | x | x | x | x | x |
| Neuropsychiatric Inventory (NPI) | x | x | x | ||
| Neuropsychiatric Inventory (NPI-Q) | x | x | x | x | x |
| Cognitive Change Index (CCI) | x | x | |||
| Everyday Cognition 12-item (ECog 12-item) | x | ||||
| Everyday Cognition 39-item (ECog) | x | x | x |
Alternative routes of administration
In some cases, instruments may be administered to participants over the phone, online, or via mail-in questionnaire, rather than in-clinic. In such cases, the alternative setting is noted in one of two places:
- A field in the corresponding table
- The VISTYPE field of the REGISTRY table
For instruments where the corresponding tables are lacking a specific field, or in cases where that field is inconsistent, entries can be cross-referenced with the VISTYPE field of the registry using the Phase, VISCODE, and RID fields to determine if a visit occurred in clinic or over the phone.
Another exception is the Cogstate Brief Battery (CBB), which carried the option of being completed digitally at home.
How instruments are scored
This varies on an instrument-by-instrument basis. Broadly speaking, there are three cases that can occur:
- The instrument is scored by staff at the clinical site, and the total is manually input into the EDC – for example, the Geriatric Depression Scale (GDS)
- Site staff input specific data elements, and the instrument is automatically scored based on that input. For example, the Clinical Dementia Rating (CDR) is scored by site staff inputting box scores into a scoring algorithm, which automatically calculates the CDR global score.
- The instrument is not scored, and only item level data are available – for example, the Assessment of Everyday Cognition (ECog)
Statistical Considerations
There are several important statistical considerations to keep in mind when analyzing data resulting from neuropsychiatric assessments, particularly at an item level:
Likert scored instruments (items are evaluated on a scale that generally ranges from “strongly agree” to “strongly disagree”) produce ordinal data, not continuous data.
Many instruments that are NOT Likert scored produce ordinal data, not continuous data. For example, global and sum-of-box scores for the Clinical Dementia Rating (CDR).
It is suboptimal, and generally inappropriate, to treat ordinal data as continuous.
Information on specific instruments/tables:
Geriatric Depression Scale (GDS)
The Geriatric Depression Scale (GDS) is a self-report measure of depression in older adults. In ADNI, the 15-item version of the scale is administered, as opposed to the longer 30-item version.
Usage Notes
Researchers who are interested in working with item-level data for the instrument should be somewhat wary.
This questionnaire consists of 15 yes/no questions, where participants score a point for ‘yes’ on 10 of the items, and score a point for ‘no’ on the remaining 5.
The CRFs used by sites to enter the results of this instrument into the EDC include a binary 1/0 for each item, as well as a score that is totaled by site staff and entered into the total score field.
Computing total scores based on item level data does not always result in the same scores that appear in the total column. Investigation into this issue is still ongoing, however it is more likely that the quality issue exists within the item level data rather than the total score column.
Montreal Cognitive Assessment (MoCA)
The Montreal Cognitive Assessment (MoCA; full version) is a widely used screening tool designed to detect mild cognitive impairment (MCI) and early dementia, including Alzheimer’s disease. It is a brief, 30-point test that assesses various cognitive domains including memory, attention, language, and visuospatial skills.
Usage Notes:
In ADNI, the MoCA is generally administered in a clinical setting. However, in cases where the participant was unable or unwilling to attend an in-person clinical visit, the full form of the instrument has also been collected via phone interview.
Note that the phone version of the instrument is not identical to the version administered in-clinic, as some of the items involve the use of visual stimuli that cannot be presented over the phone.
For data collected in the ADNI3 phase of the study, these deviations are categorized in the SOURCE field. This information was not recorded for ADNIGO and ADNI2
The total score MOCA, and the collection date MCDATE are recorded for the ADNI3 and ADNI4 phases.
Assessment of Everyday Cognition (ECOg)
The Everyday Cognition Scale (ECog) is a questionnaire designed to measure observed changes in instrumental activities of daily living that map to 6 cognitive domains. It asks participants to evaluate their ability to perform certain everyday tasks, in comparison to their ability to perform those same tasks 10 years prior. The separate, study partner report version asks study partners to evaluate the same observed changes in the participant.
There are two versions of the instrument, both of which have been used in ADNI. There is a 39-item inventory that was used in ADNIGO, ADNI2, and ADNI3, as well as a 12-item version that is currently in use for ADNI4. The 12-item version is also administered digitally as part of the ADNI4 Remote Digital cohort and Remote Longitudinal Monitoring of the in-clinic cognitively normal (CN) and mild cognitive impairment (MCI) cohorts.
The tables containing information on the 39-item ECog are named “ECOGPT” and “ECOGSP” for the participant and informant report versions of the instrument, respectively. The tables containing information on the in-clinic administered version of the 12-item ECog are titled “ECOG12PT” and “ECOG12SP”. The tables containing information on the remote administration of the 12-item ECog are titled “RMT_ECOG12PT” and “RMT_ECOG12SP”.
Usage Notes (ECog-39):
In ADNI, the ECog is generally administered in a clinical setting with both a participant and informant-version of the questionnaire. However, in cases where the participant or their study partner were unable or unwilling to attend an in-person clinical visit, results have also been collected via mail-in questionnaire or phone interview. For data collected in the ADNI3 and ADNI4 phases of the study, these deviations are categorized in the SOURCE field. This information was not recorded for ADNIGO and ADNI2
Some of the item-level data pertaining to the visuo-spatial component of the questionnaire are offset in ADNIGO and ADNI2.
- VISSPAT5 was initially duplicated and then removed in ADNIGO and ADNI2, and it should be excluded in scoring.
- Because VISSPAT5 is removed, all subsequent items are offset by 1. E.g. VISSPAT5 in ADNI3 is the same item as VISSPAT6 in ADNIGO and ADNI2.
- Because of this offset, VISSPAT8 is present only in ADNIGO and ADNI2, and is comparable to VISSPAT7 in ADNI3.
Total and component-wise scores are recorded for ADNI3 only - these need to be manually calculated for ADNIGO and ADNI2. This procedure can be carried out using the ecog.score() function in the ADNIMERGE package, or manually.
The CONCERN field, which records the participant’s response to the question “Are you concerned that you have a memory or other thinking problem?”, is not present for ADNIGO. There is also no comparable field in the study partner version of the ECog. It is also not a part of the questionnaire.
Usage Notes (ECog-12):
12-item ECog results for participants and study partners who completed the ECog as part of either the Remote Digital cohort, or the Remote Longitudinal Monitoring Cohort can be found in the “RMT_ECOG12SP” and “RMT_ECOG12PT” tables, which are located in the ‘ADNI Online’ section of the study files interface.
Mini Mental State Exam (MMSE)
The Mini-Mental State Examination (MMSE) is a brief, frequently used screening tool for Alzheimer’s disease drug studies. The MMSE scale evaluates orientation, memory, attention, concentration, naming, repetition, comprehension, and ability to create a sentence and to copy two overlapping pentagons.
Usage Notes:
Due to copyright regulations, the actual contents of the MMSE is unavailable to view without a license.
The total score for the instrument can be found in the MMSCORE field.
Clinical Dementia Rating (CDR)
The Clinical Dementia Rating is a widely used instrument that stages the severity of dementia on a scale of 0 to 5. The CDR has been administered since the beginning of the ADNI study. The CDR utilizes information gathered from an informant/study partner as well as the participant, along with the judgment of a trained test administrator.
Usage Notes:
The CDR has always used as part of the diagnostic criteria for the ADNI study, and researchers should be cautious to avoid issues in analysis that can arise from looking at the CDR and clinical diagnostic status independently, as the two are fundamentally dependent.
The CDR global score CDGLOBAL is based on a proprietary scoring algorithm, and totals CDRSB are provided in the “CDR” table.
Alzheimer’s Disease Assessment Scale Cognitive - 13 (ADAS-COG 13)
The ADAS-Cog13 is a structured scale that evaluates memory (word recall, word recognition), reasoning (following commands), language (naming, comprehension), orientation, ideational praxis (placing letter in envelope) and constructional praxis (copying geometric designs). Ratings of spoken language, language comprehension, word finding difficulty, and ability to remember test instructions are also obtained.
The ADAS-Cog13 has been collected since the beginning of the ADNI study and has historically been first in the order of tests that are administered to participants.
Usage notes
The ADAS-Cog is an example of a source of information that has been collected continuously since the beginning of the study but recorded inconsistently. There are multiple phase-specific tables that contain ADAS data.
Total scores for the ADAS-Cog across all phases have been harmonized together into the ADAS total scores table. Item-level data for historical phases (not for ADNI4) should be available in the archive section of the study files interface.
There may be some quality issues with item level data across phases, specifically related to the word lists used in the verbal recall version of the assessment. Investigation into these issues and attempts to harmonize the disparate tables are ongoing.
Neuropsychiatric Inventory (NPI) and Neuropsychiatric Inventory-Questionnaire (NPI-Q)
The NPI is a well-validated informant report instrument, developed to assess the severity of neuropsychiatric symptoms that may accompany Alzheimer’s Disease (AD) and related dementias. The NPI-Q is a shortened version of the NPI, designed for easy self-administration in shorter time settings, or over the phone.
The NPI-Q has been administered continuously since the start of ADNI, while the NPI was added in ADNI2 and has been administered since.
Usage Notes
The question of which of the two instruments is administered on a given visit - the NPI, or the shorter-form NPI-Q - has not, historically, had a consistent answer. The NPSOURC field in the “NPI” table, and the SOURCE field in the “NPIQ” table, indicate whether information was collected over the phone, or in-person. In some cases, both instruments may have been completed during the same visit.
Investigators may wish to consult the study partner information (SPINFO) table, for information on the informant demographics and relationship to the participant.
Functional Assessment Questionnaire (FAQ)
The FAQ is an informant-report questionnaire measuring impairment in activities of daily living (ADL). It has been administered continuously since ADNI1.
Usage Notes
The coding scheme for the item level data does not necessarily reflect the contribution of each item towards the total score. The scheme is explained in the figure below, using a particular item from the instrument as an example.
Cogstate Brief Battery (CBB)
The Cogstate Brief Battery is a computerized cognitive assessment meant to screen for the presence of early AD. It was administered in the ADNI2 and ADNI3 phases of the study.
Usage Notes
Information regarding the Cogstate battery are split across two tables - the results of the assessment itself can be found in the ‘Cogstate Battery Results’ table (ADNI_CBBRES), while the “CBBCOMP” table tracks the completion of the assessment in the ADNI3 phase.
Financial Capacity Instrument short form (FCI-SF)
The FCI is an assessment of financial skills in older adults, and measures both performance and time to completion on financial tasks - including mental math related to basic monetary and financial tasks, as well as the ability to interpret and use checkbooks and bank statements. The FCI was administered only in ADNI3.
Usage Notes
Note that this table contains many blank entries (all columns related to the instrument NA’d out), corresponding to cases where a visit occurred but the assessment was not conducted.
Columbia- Suicide Severity Rating Scale (C-SSRS)
This instrument measures suicidal behavior and ideation in participants, particularly as it relates to AD and cognitive decline. This instrument was first collected in ADNI4 as part of the amyloid PET disclosure process.
Usage Notes
A Spanish language version of this assessment is available, and its use is indicated in the LANGUAGE_CODE fields of the “CSSRSAD_AD” table.
Impact of Events Scale (IES)
This instrument is a self-report questionnaire that assesses subjective feelings of distress associated with particularly traumatic life events. In the ADNI study, it is used specifically in the context of amyloid PET status disclosure. This instrument was first collected in ADNI4.
Mini-State Trait Anxiety Inventory (STAI)
This is a self-report questionnaire intended to assess feelings of state and trait anxiety in participants. In the ADNI study, it is used specifically in the context of amyloid PET status disclosure. This instrument was first collected in ADNI4.
RYFF Psychological Well-Being Purpose in Life
This 14-item instrument was first collected in ADNI4 and assesses participants’ feelings of having clear goals and purpose in life. It is one component of the larger Ryff psychological well-being scale (the other components are not administered). In the ADNI study, it is used specifically in the context of the amyloid PET status disclosure process.
American National Adult Reading Test (AMNART)
The AMNART is an instrument designed to assess premorbid intelligence in older American adults. It was added to the ADNI test battery in the ADNI3 phase and has continued into ADNI4.
Usage Notes
AMNART results for ADNI4 are present in a standalone “AMNART” table, and for previous phases the information is contained in the Neuropsychological Battery table (NEUROBAT), under the ANARTND and ANARTERR fields.
More fine-grained data is available for ADNI4, while only total scores (# of errors) are recorded for previous phases (in the “NEUROBAT” table)
A Spanish language version, the Word Accentuation Test - Chicago (WATC) is available in ADNI4, in the “WATC” table.
Category Fluency Test (CFT)
The category fluency test has been used continuously since ADNI1, and is a component of the Preclinical Alzheimer’s Cognitive Composite (PACC) inventory. The test is a measure of semantic fluency, which requires participants to generate examples of animals.
A version of the instrument that required participants to generate examples of vegetables was used in ADNI1, and dropped in subsequent phases.
Usage Notes
Data from the Animals version of the assessment can be found in the Neuropsychological Battery table (NEUROBAT) in the CATANIMSC, CATANPERS, and CATANINTR fields.
Data from the vegetables version of the assessment (ADNI1 only) can also be found in the NEUROBAT table, in the CATVEGESC, CATVGPERS, and CATVGINTR fields.
Logical Memory: Delayed and Immediate Recall (LM)
These tests are both components of the Wechsler Memory Scale. Individuals are read a story, and are prompted to recall details both immediately, and after a period of time has elapsed. The instruments are evaluated based on the veracity of the recalled information, and the need for cues from the clinician administering the test.
Usage Notes
Information on both assessments is contained in the Neuropsychological Battery (NEUROBAT) table. The immediate recall portion can be found in the LMSTORY, LIMMTOTAL, and LIMMEND fields, while the delayed recall portion is found in the LDELBEGIN, LDELTOTAL, and LDELCUE fields.
Rey Auditory Verbal Learning Test (RAVLT)
This is an assessment of memory and verbal learning in which participants are given a list of unrelated words and prompted to recite items from that list both immediately, and after a 5-25 minute period of time has elapsed.
Usage Notes
Results for both the immediate and delayed recall portions of the RAVLT can be found in the Neuropsychological Battery table (NEUROBAT). Results of the immediate recall portion are found in the AVTOT1, AVERR1, AVTOT2, AVERR2, AVTOT3, AVERR3, AVTOT4, AVERR4, AVTOT5, AVERR5, AVTOT6, AVERR6, AVTOTB, AVERRB, and AVENDED fields.
Results of the delayed recall portion of the assessment are found in the AVDELBEGAN, AVDEL30MIN, AVDELERR1, AVDELTOT, and AVDELERR2 fields.
Further information can be found in the RAVLT.learning, RAVLT.immediate, RAVLT.forgetting, and RAVLT.perc.forgetting fields.
Digit Span Test
The Digit Span test requires participants to memorize and recall a list of digits, sometimes in different orders to that in which the digits were initially presented. This instrument was only collected during the ADNI1 phase.
Usage Notes
The results of this assessment are recorded in the Neuropsychological Battery (NEUROBAT) table, in the DSPANFOR, DSPANFLTH, DSPANBAC, and DSPANBLTH fields.
Clock Drawing Test (CDT) and Clock Copying Test
The CDT is used to evaluate visuospatial abilities in participants. In this assessment, participants are either given instructions to draw a clock and indicate a specific time or given an example of a clock to copy.
Usage Notes
Results for this assessment are found in the Neuropsychological Battery table (NEUROBAT). The results of the free-hand clock drawing test are contained in the CLOCKCIRC, CLOCKSYM, CLOCKNUM, CLOCKHAND, CLOCKTIME, and CLOCKSCOR fields. The results of the clock copying test are found in the COPYCIRC, COPYSYM, COPYNUM, COPYHAND, COPYTIME, and COPYSCOR fields.
Trail making test (TMT)
The trail making test is an assessment of multiple cognitive domains, which requires participants to quickly draw lines connecting circles based on information contained within those circles. It is commonly used as a component of different Preclinical Alzheimer’s Cognitive Composite (PACC) scores.
Usage Notes
The results of this assessment are found in the Neuropsychological Battery table (NEUROBAT), in the TRAASCOR, TRAAERRCOM, TRAAERROM, TRABSCOR, TRABERRCOM, and TRABERROM fields.
Boston Naming Test (BNT) and Mulitlingual Naming test (MINT)
The BNT requires participants to identify objects based on crude line drawings. A Spanish-language version of the assessment has been available since ADNI-GO. The BNT was replaced with a similar instrument, the MiNT, for the ADNI3 and ADNI4 phases.
Usage Notes
Results for both assessments can be found in the Neuropsychological Battery table (NEUROBAT). Information on the BNT is captured in the BNTND, BNTSPONT, BNTSTIM, BNTCSTIM, BNTPHON, BNTCPHON, and BNTTOTAL fields. Information on the MiNT is captured in the MINTUNCUED, MINTSEMCUE, and MINTTOTAL fields.