ADNI-D

The ADNI Depression Project (ADNI-D) is a companion study to the Alzheimer's Disease Neuroimaging Initiative that asks why older adults with major depression show cognitive impairment and faster cognitive decline. ADNI itself excluded people with major depression, so ADNI-D enrolled them separately and imaged them with ADNI methods, so that brain atrophy, white matter lesions, cerebral blood flow and amyloid burden can be compared with non-depressed ADNI participants. The project started in 2013 under R. Scott Mackin at UCSF and is funded by NIMH with a contribution from Eli Lilly. Data are distributed by LONI through a separate ADNI-D database on the Image and Data Archive.

Composition

The protocol planned 120 participants aged 65 or older with a current DSM-IV diagnosis of unipolar major depression without psychosis, an episode of at least six weeks and a Hamilton depression score of 15 or more, while dementia was excluded. Non-depressed comparison data are taken from ADNI rather than collected in ADNI-D. The first main paper analysed 119 depressed participants with a mean age of 70.9 years (range 65 to 91), 80 of them women; 39 met ADNI criteria for mild cognitive impairment.

Acquisition

Participants were recruited through the psychiatry departments of UCSF and the University of Pittsburgh and scanned at the matching ADNI sites, which had taken part in ADNI-2. The baseline visit includes a 3T MRI with an accelerated 3D T1-weighted scan, axial FLAIR, T2* gradient echo, DTI, arterial spin labeling, functional MRI, a fat-saturated T2 fast spin echo and a field map, plus florbetapir amyloid PET acquired as four 5-minute frames from 50 minutes after injection. Clinical, cognitive and mood assessments, blood for APOE genotyping, GWAS and plasma biomarkers are collected at baseline and again after about 30 months, with telephone checks at 12 and 24 months.

Annotations

There are no image labels. Each participant has a SCID diagnosis, a depression and treatment history, depression rating scales and the ADNI cognitive battery. Radiologists give a local clinical read of every MRI.

Known limitations

The cohort is small and comes from two sites, while ADNI comparison participants were scanned at many sites and scanner types. The total number enrolled and imaged is not stated in the public sources found. The protocol schedules MRI and PET at baseline only, so imaging is cross-sectional.