Brain structure in bipolar disorder: A longitudinal neuroimaging study in twins

In this longitudinal twin study, the goal was to ascertain the extent to which BD shows abnormalities in subcortical and cortical brain regions at baseline and over time. Baseline assessments in this study revealed that BD was associated with smaller volumes of the thalamus, putamen and nucleus accumbens. These associations were strongly influenced by genes influencing both BD and smaller volumes of these regions, indicating that BD may share some genetic background with structural abnormalities in brain regions involved in emotion processing. Furthermore, at baseline, BD was associated with subtle deficits in cortical surface area, cortical thickness and cortical volume. Here, unique environmental factors primarily drove the associations between BD and cortical measures, particularly those involving cortical thinning. In contrast, although brain structure deficits may arise during development, these appear to remain relatively stable over time in adulthood, as structural brain changes in cortical and subcortical brain regions were not different among BD patients, co-twins and healthy controls. This suggests that BD is not a disease that shows progressive structural brain atrophy and may be specifically associated with subtle brain abnormalities that arise earlier in life, perhaps even during neurodevelopment. This finding contrasts with results obtained in schizophrenia, where significant progressive brain atrophy during adulthood, particularly in fronto-temporal regions, has been observed and linked to the genetic liability for the disease. Therefore, schizophrenia and BD may show different neuroanatomical trajectories, with the former showing more severe abnormalities than the latter. However, more longitudinal studies with larger cohorts than the one described here are necessary to confirm the absence of structural brain change over time in BD, especially since findings of structural brain changes in prefrontal cortical and subcortical structures conflict considerably between studies. In addition, many factors may affect brain measures and could contribute to inconsistencies across studies, such as lithium use, age, familial load, mood status, number of episodes, number of hospitalizations and variability in imaging methodology.

Finally, to account more comprehensively for the genetic and environmental contributions to structural brain abnormalities in BD, cross-sectionally and longitudinally, (prospective) longitudinal studies with larger cohorts (such as those included by the ENIGMA consortium) are required. More insight into which specific genetic and/or environmental influences contribute to the development of BD and associated brain anomalies, particularly in brain networks subserving emotion processing, could serve as a guide in the developing intervention strategies directed at reducing disease burden or even preventing disease onset.