Searching high and low for meaningful and replicable morphometric correlates of personality.

Abstract

Recent personality neuroscience research in large samples suggests that personality traits tend to bear nullto-small relations to morphometric (i.e., brain structure) regions of interest (ROIs). In this preregistered, two-part study using Human Connectome Project data (N = 1,105), we address the possibility that these null-to-small relations are due, in part, to the "level" (i.e., hierarchical placement) of personality and/or morphometry examined. We used a Five-Factor Model framework and operationalized personality in terms of meta-traits, domains, facets, and items; we operationalized morphometry in terms of omnibus measures (e.g., total brain volume), and cortical thickness and area in the ROIs of the Desikan and Destrieux atlases. First, we compared the patterns of effect sizes observed between these levels using mixed effects modeling. Second, we used a machine learning framework for estimating out-of-sample predictability. Results highlight that personality-morphometry relations are generally null-to-small no matter how they are operationalized. Relatively, the largest mean effect sizes were observed at the domain level of personality, but the largest individual effect sizes were observed at the facet and item level, particularly for the Ideas facet of Openness and its constituent items. The largest effect sizes observed were at the omnibus level of morphometry, and predictive models containing only omnibus variables were comparably predictive to models including both omnibus variable and ROIs. We conclude by encouraging researchers to search across levels of analysis when investigating relations between personality and morphometry and consider prioritizing omnibus measures, which appear to yield the largest and most consistent effects.

Structural brain differences do not mediate the relations between sex and personality or psychopathology

“…Specifically, we test if brain morphometry statistically mediates the relations between sex and FFM neuroticism and agreeableness, as well as between sex and internalizing and externalizing psychopathology using data from the Human Connectome Project (HCP; Van Essen et al, 2012. Given the findings from the meta-analysis by Eliot and colleagues (2021), we prioritized omnibus brain variables (i.e., total brain volume, total gray matter volume, total white matter volume, total cortical surface area, mean cortical thickness) as the primary mediators of interest, which is consistent with studies that have linked neuroticism to total brain volume (Bjørnebekk et al, 2013;Jackson et al, 2011;Liu et al, 2013), including work from our own team in this sample (Hyatt et al, 2021). Next, we conducted two sets of secondary analyses using regional morphometric indices defined by different atlases as mediators.…”

Taking an elemental approach to the conceptualization and measurement of Machiavellianism, narcissism, and psychopathy

“…It also tells us nothing about life experiences that may be associated with the emergence of psychopathic behaviors. Such questions are better addressed by using a neurobiological (e.g., Kiehl, 2006; but see also; Hyatt et al., 2021) or developmental (e.g., Jonason et al., 2014) approach. In other words, the best approach to a question about personality depends on the type of question.…”

Revising the trait model of the alternative model of personality disorders: Comment on Clark and Watson’s structural review.

“…Clark and Watson (2022) also indicate that they did not consider neurobiomarkers in their proposed revisions due to a lack of research in the field; again, the lack of research here is no fault of their own. Although we agree that there is a dearth of research on the AMPD traits (see, however, Hyatt et al, 2021), we believe that some consideration be given to the fact that some of these phenotypic traits in the AMPD may have potentially unique neurobiological (or cognitive/behavioral) markers. To this end, we think it important to examine the association between the phenotypic traits included in the AMPD and their possible association with neurobiological markers.…”