• bipolar disorder;
  • late-onset mood disorders;
  • magnetic resonance imaging;
  • vascular risk factors;
  • white matter hyperintensities

Objectives:  Magnetic resonance imaging (MRI) studies have reported an increased frequency of white matter hyperintensities (WMH) in association with late-onset (LO) depression, and this has supported the notion that vascular-related mechanisms may be implicated in the pathophysiology of LO mood disorders. Recent clinical studies have also suggested a link between LO bipolar disorder (LO-BD) and cerebrovascular risk factors, but this has been little investigated with neuroimaging techniques. In order to ascertain whether there could be a specific association between WMH and LO-BD, we directly compared WMH rates between LO-BD subjects (illness onset ≥ 60 years), early-onset BD subjects (EO-BD, illness onset <60 years), and elderly healthy volunteers.

Methods:  T2-weighted MRI data were acquired in LO-BD subjects (n = 10, age = 73.60 ± 4.09), EO-BD patients (n = 49, age = 67.78 ± 4.44), and healthy subjects (n = 24, age = 69.00 ± 7.22). WMH rates were assessed using the Scheltens scale.

Results:  There was a greater prevalence of WMH in LO-BD patients relative to the two other groups in the deep parietal region (p = 0.018) and basal ganglia (p < 0.045). When between-group comparisons of mean WMH scores were conducted taking account of age differences (ANCOVA), there were more severe scores in LO-BD patients relative to the two other groups in deep frontal and parietal regions, as well as in the putamen (p < 0.05).

Conclusions:  Our results provide empirical support to the proposed link between vascular risk factors and LO-BD. If extended in future studies with larger samples, these findings may help to clarify the pathophysiological distinctions between bipolar disorder emerging at early and late stages of life.