Identifying subtypes of bipolar disorder based on clinical and neurobiological characteristics

Abstract

The ability to classify patients with bipolar disorder (BD) is restricted by their heterogeneity, which limits the understanding of their neuropathology. Therefore, we aimed to investigate clinically discernible and neurobiologically distinguishable BD subtypes. T1-weighted and resting-state functional magnetic resonance images of 112 patients with BD were obtained, and patients were segregated according to diagnostic subtype (i.e., types I and II) and clinical patterns, including the number of episodes and hospitalizations and history of suicide and psychosis. For each clinical pattern, fewer and more occurrences subgroups and types I and II were classified through nested cross-validation for robust performance, with minimum redundancy and maximum relevance, in feature selection. To assess the proportion of variance in cognitive performance explained by the neurobiological markers, multiple linear regression between verbal memory and the selected features was conducted. Satisfactory performance (mean accuracy, 73.60%) in classifying patients with a high or low number of episodes was attained through functional connectivity, mostly from default-mode and motor networks. Moreover, these neurobiological markers explained 62% of the variance in verbal memory. The number of episodes is a potentially critical aspect of the neuropathology of BD. Neurobiological markers can help identify BD neuroprogression.

Figure 1

Number of features with different thresholds.

Figure 2

The major features of the classification analyses for the clinical patterns. After 100 processes of minimum redundancy maximum relevance selection with the features selected through with the threshold of the most optimal mean performance, the major features were selected over 90 times during the outer loop of the nested cross-validation. The illustration uses eight networks of Shen’s 268-region parcellation. The red lines represent the connectivity of the more occurrences group being higher than that of the fewer occurrences group, and the blue lines represent the opposite direction. (a) The major features of the classification for the number of episodes. (b) The major features of the classification based on the number of hospitalizations. (c) The major features of the classification based on suicide attempts. (d) The major features of the classification based on the history of psychosis. (e) The major features of the classification based on diagnostic subtypes. The red lines represent the connectivity of BDI being higher than that of BDII, and the blue lines represent the opposite direction.

Figure 3

The involved regions of major features of the classification analyses for fewer and more occurrences groups based on the number of episodes, the number of hospitalizations, suicide attempts, and having psychosis or not, which were selected more than 90 times after 100 times of processing of minimum redundancy maximum relevance selection during the outer loop of the nested cross-validation, as illustrated on the glass brain. (a) The involved regions of major features of the classification for the number of episodes. (b) The involved regions of major features of the classification based on the number of hospitalizations. (c) The involved regions of major features of the classification based on suicide attempts. (d) The involved regions of major features of the classification based on the history of psychosis. (e) The involved regions of major features of the classification based on diagnostic subtypes. (ACC anterior cingular cortex, InfO inferior occipital cortex, InfOperF inferior opercular frontal cortex, InfOrbF inferior orbitofrontal cortex, InfT inferior temporal cortex, InfTriF inferior triangular frontal cortex, MCC middle cingular cortex, MidF middle frontal cortex, MidO middle occipital cortex, MidTP middle temporal pole, ParaCen paracentral lobule, PCC posterior cingular cortex, PostCen postcentral cortex, PreCen precentral cortex, SMA supplementary motor area, SupF superior frontal cortex, SupMedF superior medial frontal cortex, SupO superior occipital cortex, SupOrbF superior orbitofrontal cortex, SupP superior parietal cortex, SupT superior temporal cortex).