Scottish Bipolar Family Study fMRI (Whalley 2013)

This entry covers the baseline functional MRI sample of the Scottish Bipolar Family Study, run by the Division of Psychiatry at the University of Edinburgh. Psychiatrists across Scotland identified patients with bipolar I disorder, who named relatives aged 16 to 25. Relatives with no mood or psychotic illness themselves were scanned once and reassessed about two years later. Whalley and colleagues (PLOS ONE, 2013) used the sample to ask whether brain activation at baseline separates the relatives who went on to develop major depression from those who stayed well. They reported higher insula activation with rising task difficulty in those who later became ill.

Composition

Baseline scans exist for 110 high-risk relatives (at least one first-degree or two second-degree relatives with bipolar I disorder) and 70 unrelated controls drawn from the relatives' social networks and matched for age, sex and premorbid IQ. Follow-up status could not be established for 10 relatives and 8 controls, and 4 controls who developed depression were dropped. The analysed sample is 156: 58 controls, 78 relatives who stayed well and 20 who developed major depression. Two further relatives developed bipolar disorder. Mean age per group is about 21 years. Clinical measures include the Hamilton depression and Young mania scales, TEMPS-A temperament and the NEO five-factor inventory.

Acquisition

All scans were acquired on a GE 1.5 T Signa scanner at the Scottish Brain Imaging Research Centre in Edinburgh. The functional run was gradient-echo EPI (TR 2 s, TE 40 ms, 27 axial 5 mm slices, 404 volumes) during a block-design Hayling sentence completion task with four difficulty levels. A T1-weighted volume with 180 coronal 1.2 mm slices served for registration. Processing used SPM5.

Annotations

No image annotations. Labels are group membership, follow-up diagnosis from structured interview or health records, and the clinical and personality scores.

Known limitations

• The scans have no public release and the paper gives no data access route or terms.
• The figshare record holds supplementary documents only, not imaging data.
• Sex counts cover the 156 analysed subjects, not all 180 scanned.
• Follow-up for 33 subjects came from written contact with the health service rather than an interview.