Diagnostic Information
Some of the most frequent questions received from ADNI data users are regarding the diagnostic status. Tables containing diagnostic information can be found on the IDA under the Assessments -> Diagnosis subheading.
There are three tables that contain diagnostic information: baseline changes/clinical status (BLCHANGE), diagnostic and symptoms checklist (ADSXLIST), and diagnostic summary (DXSUM).
Out of these three tables, the diagnostic summary (DXSUM) provides a straightforward clinical diagnosis and is the table that should be used for most applications.
BLCHANGE offers some more granular diagnostic information, and ADSXLIST captures ancillary non-cognitive symptoms presented by participants and was only used in the ADNI1 and ADNIGO phases of the study.
For more specific information on diagnostic criteria, please consult the section on [screening visits and diagnostic cohorts].
Baseline changes/clinical status (BLCHANGE)
In ADNI4, this table is referred to as ‘diagnostic summary clinical status’, while in previous phases it was called ‘diagnostic summary baseline changes’.
This table captures information on any clinically relevant changes in the participant’s cognitive status, where applicable. Note that, in this table, missing data are indicated with -4 in older phases and coded as NA in newer phases.
Diagnostic and symptoms checklist (ADSXLIST)
This table includes other participant-reported symptoms that may be diagnostically relevant, including headache, depressed mood, etc. It was only collected in the ADNI1 and ADNIGO phases of the study.
Diagnostic Summary (DXSUM)
This is the primary table for clinical diagnostic information. It captures information on a participant’s diagnostic cohort, and in some cases more granular information regarding participants who are assigned into an impaired cohort.
DIAGNOSIS is the field that captures a participant’s clinical diagnosis. For any given participant, and any given time point, this field provides their clinically assessed diagnosis at that time point.
The coding scheme is:
| Value | Meaning |
|---|---|
| 1 | Cognitively Unimpaired |
| 2 | MCI |
| 3 | Dementia |
Entries in the data dictionary that indicate otherwise are out of date, as the field was harmonized towards the beginning of the ADNI4 phase.
The dictionary entry for DIAGNOSIS in the ADNI2 phase reflects a different field that is no longer in use.
DXCURREN and DXCHANGE are older fields that are no longer in use. They were the primary source for clinical diagnostic information in the ADNI1,GO, and 2 phases, but they were later translated into the DIAGNOSIS field.
Notably, the DXCHANGE field coded diagnostic status relative to the diagnosis assigned during the previous visit. Data users who prefer this scheme can recreate it by comparing the DIAGNOSIS field with its value at a previous timepoint.
The DXCONFID field indicates clinician confidence in their diagnostic assessment.
Several fields in the table offer a more granular characterization of impairment.
DXPARK indicates the presence of symptoms associated with parkinsonism, and for the ADNI1 phase specifically there are additional fields that give further characterization of those symptoms if they are present.
For participants assigned a diagnosis of MCI, DXMDES distinguishes between amnesiac and non-amnesiac MCI while DXMDUE gives a clinician’s assessment of whether the MCI presented by the participant can be attributed to AD, or to some other etiology.
The six DXMPTR fields (DXMPTR1, DXMPTR2, ..., DXMPTR6) further characterize the clinical features of impairment for participants with a diagnosis of MCI.
For participants in the dementia cohort DXAPP, DXAPROB, and DXAPOSS characterize the clinician’s confidence in assigning a diagnosis of AD, rather than dementia due to some other etiology.
The DXOTHDEM and DXOTHDES fields further expand on the suspected etiology of dementia for participants who exhibit symptoms that do not match the profiles of either AD or Parkinsonism (With the former being exclusive to the ADNI1 phase).