Background: Validating the high-risk (HR) and ultra-high-risk (UHR) stages of bipolar disorder (BP) may help enable early intervention strategies.
Methods: We followed up with 44 offspring of parents with BP, subdividing into the HR and UHR categories. The offspring were aged 8-28 years and were free of any current DSM-IV diagnoses. Our multilevel, integrative approach encompassed gray matter (GM) volumes, brain network connectivity, neuropsychological performance, and clinical outcomes.
Findings: Compared with the healthy controls (HCs) (n = 33), the HR offspring (n = 26) showed GM volume reductions in the right orbitofrontal cortex. Compared with the HR offspring, the UHR offspring (n = 18) exhibited increased GM volumes in four regions. Both the HR and UHR offspring displayed abnormalities in the inferior occipital cortex regarding the measures of degree and centrality, reflecting the connections and roles of the region, respectively. In the UHR versus the HR offspring, the UHR offspring exhibited upwards-shifted small world topologies that reflect high clustering and efficiency in the brain networks. Compared with the HCs, the UHR offspring had significantly lower assortativity, which was suggestive of vulnerability. Finally, processing speed, visual-spatial, and general function were impaired in the UHR offspring but not in the HR offspring.
Interpretation: The abnormalities observed in the HR offspring appear to be inherited, whereas those associated with the UHR offspring represent stage-specific changes predisposing them to developing the disorder.
Keywords: Affective disorder; Bipolar disorder; Cognition; High-risk design; Network analysis; Neuroimaging; Ultra-high-risk.