Elsevier

NeuroImage

Volume 174, 1 July 2018, Pages 164-176
NeuroImage

Neural and genetic determinants of creativity

https://doi.org/10.1016/j.neuroimage.2018.02.067Get rights and content

Highlights

  • The neural correlates and genetic determinates of creativity remain largely unclear.

  • We developed a prediction-based approach in identifying the functional connectivity and SNPs closely related to creativity.

  • High figural creativity is associated with brain networks with strong top-down control versus weak bottom-up processes.

  • Genes correlated with figural creativity were involved in glutamate and GABA functionality.

  • Our neuroimaging prediction model was cross-validated by a completely new dataset.

Abstract

Creative thinking plays a vital role in almost all aspects of human life. However, little is known about the neural and genetic mechanisms underlying creative thinking. Based on a cross-validation based predictive framework, we searched from the whole-brain connectome (34,716 functional connectivities) and whole genome data (309,996 SNPs) in two datasets (all collected by Southwest University, Chongqing) consisting of altogether 236 subjects, for a better understanding of the brain and genetic underpinning of creativity. Using the Torrance Tests of Creative Thinking score, we found that high figural creativity is mainly related to high functional connectivity between the executive control, attention, and memory retrieval networks (strong top-down effects); and to low functional connectivity between the default mode network, the ventral attention network, and the subcortical and primary sensory networks (weak bottom-up processing) in the first dataset (consisting of 138 subjects). High creativity also correlates significantly with mutations of genes coding for both excitatory and inhibitory neurotransmitters. Combining the brain connectome and the genomic data we can predict individuals' creativity scores with an accuracy of 78.4%, which is significantly better than prediction using single modality data (gene or functional connectivity), indicating the importance of combining multi-modality data. Our neuroimaging prediction model built upon the first dataset was cross-validated by a completely new dataset of 98 subjects (r = 0.267, p = 0.0078) with an accuracy of 64.6%. In addition, the creativity–related functional connectivity network we identified in the first dataset was still significantly correlated with the creativity score in the new dataset (p<103). In summary, our research demonstrates that strong top-down control versus weak bottom-up processes underlie creativity, which is modulated by competition between the glutamate and GABA neurotransmitter systems. Our work provides the first insights into both the neural and the genetic bases of creativity.

Introduction

Creativity in humans is a complex cognitive behavior commonly defined as the ability to generate responses that are novel, useful, and appropriate, (Sternberg and Lubart, 1996). Creative thinking plays an important role in almost all aspects of our life, most prominently in the arts, science, and engineering.

Divergent thinking tests are so far the major type of psychometric instrument in creativity testing (Gardner, 1988; Plucker and Runco, 1998), including various verbal and figural tasks. Based on the divergent thinking tests, mounting evidence suggests that the default mode network and brain regions associated with cognitive control may be important for creativity. For example, regions of the default mode network, including the precuneus (Fink et al., 2014a, 2014b; Jauk et al., 2015; Jung et al., 2010; Takeuchi et al., 2010) and inferior parietal lobule (Takeuchi et al., 2011), and executive control network (ECN) (Ellamil et al., 2012; Gonen-Yaacovi et al., 2013) have been implicated in neuroimaging studies of divergent thinking tasks. The activation of the default mode network (DMN) in creative processes may reflect the spontaneous generation of candidate ideas, and/or the retrieval of long-term memory (Beaty et al., 2016), while the control network may serve to constrain cognition to meet specific goals of the tasks.

As a highly complex cognitive process, creativity relies on not only the activity of separate brain regions but also the interactions between different brain regions/networks (Bullmore and Sporns, 2009; Fox et al., 2005; van den Heuvel and Hulshoff Pol, 2010). Functional connectivity (FC) (the correlation between the BOLD signal of different brain regions) depicts the interaction between different brain regions, which is informative for complex behaviors such as creativity and cognition. For instance, Beaty et al. (2014) found greater connectivity between the entire DMN and the left inferior frontal gyrus (IFG) within the ECN in highly creative individuals. In addition, a recent study found that the strength of connectivity between the DMN and the frontoparietal network (FPN) was positively related to both visual and verbal creativity using independent component analysis (Zhu et al., 2017). Such findings suggest that creative thought may benefit from the cooperation of the DMN and ECN.

The genetic basis of divergent thinking has been previously explored mainly by using candidate gene approaches. Several brain regions in the dopaminergic (DA) system have been found to be involved in creativity cognition (Flaherty, 2005; Heilman et al., 2003; Takeuchi et al., 2010). Thus, many genes in the dopamine pathway, like D2 Dopamine Receptor (DRD2) (de Manzano et al., 2010; Reuter et al., 2006), D4 Dopamine Receptor (DRD4) (Mayseless et al., 2013), Tryptophan Hydroxylase (TPH1) (Reuter et al., 2006), Dopamine Transporter 1 (DAT1), and Catechol-O-Methyltransferase (COMT) (Zabelina et al., 2016), were selected as candidate genes and were found to be associated with divergent thinking. In addition, as several psychiatric illnesses are found to have genetic links to creativity (Kyaga et al., 2011; Power et al., 2015), some genes involved in the pathology of psychosis, such as neuregulin 1 (NRG1), were selected and were also found to be associated with creativity (Keri, 2009).

Although previous studies have explored the relationship between brain networks and creativity, most studies have focused on the regions previously implicated in creativity (ROI-based), or used correlation analysis, i.e., correlation between neuroimaging metrics and behavioral scores, to search for the neural basis of creativity (correlation-based). One the one hand, the ROI-based analysis may limit the potential search to only a small part of the whole brain, thus preclude the possibility of new findings (Xia and He, 2017). On the other hand, correlation-based approaches have the disadvantage of overfitting the data and often fail to generalize to novel data (Shen et al., 2017). In addition, although several studies have explored creativity-related candidate genes, a search for the genetic basis of creativity at a genome-wide scale, especially for divergent thinking, has not been described as far as we know. The correlation-based methods have also been widely employed in the genetic analysis of divergent thinking which have the disadvantage of low generalization ability.

Based on all the above, in the present study we used a large sample of 236 individuals (which we separated into two non-overlapping datasets) in this research, which enabled statistical analyses of whole-brain FCs (34,716 FCs) and whole-genome genes (309,996 SNPs) that were related to creativity without a priori hypotheses, therefore not creating constraints on the brain regions and genes to be investigated. Furthermore, we used a cross-validation based approach to identify the functional brain networks that are predictive of creativity, to ensure that the results were statistically firm and could be generalized to new populations. Our approach, IMuDP (Integrating Multi-scale Data for Prediction), is improved from one that has been used to predict sustained attention in resting-state fMRI (Rosenberg et al., 2016), which can now deal with multi-modal data. The prediction-based approach is held to provide fMRI-derived statistics that relate to individual behaviors with more generalizability than traditional correlation-based analysis (Dubois and Adolphs, 2016). Furthermore, to understand the genetic underpinning of creative thinking, we also searched from the whole genome the SNPs (single nucleotide polymorphisms) whose mutations correlate most closely with the creativity score using the same prediction-based strategy. Part of the aim here was to investigate how multi-modal data (in this case fMRI and genetic data) can be combined, and whether this helps better predictions to be made. This research provides the first whole-brain functional network analysis for creativity with a corresponding whole-genome search for related SNPs, and is aimed to shed light on both the neural and genetic underpinnings of creative thinking.

Section snippets

Ethics statement

Both the behavioral and MRI protocols were approved by the local ethics committee of Southwest China University, Chongqing. Written informed consent was obtained from all participants prior to the study, which was approved by the Institutional Human Participants Review Board of Southwest University Imaging Center for Brain Research. The methods were conducted in accordance with approved guidelines.

Participants

Three hundred and fifteen subjects were recruited in this study (Li et al., 2016). We firstly

FMRI data preprocessing

All fMRI data were preprocessed using SPM8 (http://www.fil.ion.ucl.ac.uk/spm) and the Data Processing Assistant for Resting-State fMRI (DPARSF) (Chao-Gan and Yu-Feng, 2010). We first discarded the first 10 EPI scans to suppress the equilibration effects and the remaining scans were slice timing corrected, realigned and normalized to a standard template (Montreal Neurological Institute) based on T1 images using linear transformation and resampled to 3 × 3 × 3 mm3. Next, spatial smoothing (8 mm

Prediction of the TTCT figural score

Using the whole brain functional connectome, the correlation between the TTCT score predicted by the optimal prediction model with the real TTCT score was 0.424 (p = 2.19×107, Fig. 2a). With whole genome data, the correlation between the predicted and real TTCT score was 0.466 (p = 8.70×109, Fig. 2a). The mean absolute percentage errors (MAPE) of the two prediction models were 23.9% and 22.5%, respectively (76.1% and 77.5% in terms of accuracy). With combined neuroimaging and genomic data,

Discussion

In this work, we use a cross-validation based predictive framework to search for the imaging and genetic correlates of creativity to provide more generalization than traditional correlation analyses. We found that models using the network strength of the creativity network or the mutation strength of the creativity polygenic alliance could predict the individual's performance with high accuracy, indicating that markers for creativity are present in both the FC between brain regions and in the

Funding

J. Feng is partially supported by the key project of Shanghai Science & Technology Innovation Plan (No. 15JC1400101 and No. 16JC1420402) and the National Natural Science Foundation of China (Grant No. 71661167002 and No. 91630314). There search was also partially supported by the Shanghai AI Platform for Diagnosis and Treatment of Brain Diseases (No. 2016-17). The research was also partially supported by Base for Introducing Talents of Discipline to Universities (No. B18015). Z. Liu and JY.

Acknowledgements

We would like to thank Prof. Keith M. Kendrick, Prof. Fengzhu Sun, and Jessie Liu for their helpful suggestions.

References (82)

  • K.M. Mihov et al.

    Hemispheric specialization and creative thinking: a meta-analytic review of lateralization of creativity

    Brain Cogn.

    (2010)
  • S. Naisbitt et al.

    Shank, a novel family of postsynaptic density proteins that binds to the NMDA receptor/PSD-95/GKAP complex and cortactin

    Neuron

    (1999)
  • J.D. Power et al.

    Functional network organization of the human brain

    Neuron

    (2011)
  • J.D. Power et al.

    Methods to detect, characterize, and remove motion artifact in resting state fMRI

    Neuroimage

    (2014)
  • M. Reuter et al.

    Identification of first candidate genes for creativity: a pilot study

    Brain Res.

    (2006)
  • E.T. Rolls

    A non-reward attractor theory of depression

    Neurosci. Biobehav Rev.

    (2016)
  • H. Takeuchi et al.

    Failing to deactivate: the association between brain activity during a working memory task and creativity

    Neuroimage

    (2011)
  • H. Takeuchi et al.

    Regional gray matter volume of dopaminergic system associate with creativity: evidence from voxel-based morphometry

    Neuroimage

    (2010)
  • C.A. Turner et al.

    The fibroblast growth factor family: neuromodulation of affective behavior

    Neuron

    (2012)
  • M.P. Van den Heuvel et al.

    Exploring the brain network: a review on resting-state fMRI functional connectivity

    Eur. Neuropsychopharmacol.

    (2010)
  • D. Wei et al.

    Increased resting functional connectivity of the medial prefrontal cortex in creativity by means of cognitive stimulation

    Cortex

    (2014)
  • M. Xia et al.

    Functional connectomics from a “big data” perspective

    Neuroimage

    (2017)
  • C.G. Yan et al.

    A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics

    Neuroimage

    (2013)
  • X.N. Zuo et al.

    Toward reliable characterization of functional homogeneity in the human brain: preprocessing, scan duration, imaging resolution and computational space

    Neuroimage

    (2013)
  • B.P. Ander et al.

    Atypical miRNA expression in temporal cortex associated with dysregulation of immune, cell cycle, and other pathways in autism spectrum disorders

    Mol. Autism

    (2015)
  • J.R. Andrews-Hanna et al.

    The default network and self-generated thought: component processes, dynamic control, and clinical relevance

    Ann. N. Y. Acad. Sci.

    (2014)
  • C. Best et al.

    The relationship between subthreshold autistic traits, ambiguous figure perception and divergent thinking

    J. Autism Dev. Disord.

    (2015)
  • J. Bhattacharya et al.

    Drawing on mind's canvas: differences in cortical integration patterns between artists and non-artists

    Hum. Brain Mapp.

    (2005)
  • B. Biswal et al.

    Functional connectivity in the motor cortex of resting human brain using echo-planar MRI

    Magn. Reson Med.

    (1995)
  • S. Blunt

    The creative brain: Fundamental features, associated conditions and unifying neural mechanisms

    Neurol. music

    (2010)
  • M. Boccia et al.

    Where do bright ideas occur in our brain? Meta-analytic evidence from neuroimaging studies of domain-specific creativity

    Front. Psychol.

    (2015)
  • E.M. Bowden et al.

    Aha! Insight experience correlates with solution activation in the right hemisphere

    Psychon. Bull. Rev.

    (2003)
  • E. Bullmore et al.

    Complex brain networks: graph theoretical analysis of structural and functional systems

    Nat. Rev. Neurosci.

    (2009)
  • J. Buxbaum et al.

    Association between a GABRB3 polymorphism and autism

    Mol. psychiatry

    (2002)
  • D.T. Campbell

    Blind variation and selective retentions in creative thought as in other knowledge processes

    Psychol. Rev.

    (1960)
  • Y. Chao-Gan et al.

    DPARSF: a MATLAB toolbox for “pipeline” data analysis of resting-state fMRI

    Front. Syst. Neurosci.

    (2010)
  • C. Correia et al.

    Association of the alpha4 integrin subunit gene (ITGA4) with autism

    Am. J. Med. Genet. B Neuropsychiatr. Genet.

    (2009)
  • O. de Manzano et al.

    Thinking outside a less intact box: thalamic dopamine D2 receptor densities are negatively related to psychometric creativity in healthy individuals

    PLoS One

    (2010)
  • A. Dietrich

    The cognitive neuroscience of creativity

    Psychon. Bull. Rev.

    (2004)
  • A. Dietrich et al.

    A review of EEG, ERP, and neuroimaging studies of creativity and insight

    Psychol. Bull.

    (2010)
  • D.J. Falcone et al.

    A modified lateral eye-movement measure, the right hemisphere and creativity

    Percept. Mot. Ski.

    (1984)
  • Cited by (56)

    • Brain networks of creative cognition

      2023, Handbook of Organizational Creativity: Individual and Group Level Influences, Second Edition
    • Creative problem solving in knowledge-rich contexts

      2022, Trends in Cognitive Sciences
    • The neurobiological basis of divergent thinking: Insight from gene co-expression network-based analysis

      2021, NeuroImage
      Citation Excerpt :

      To date, there have been only a few published studies trying to integrate genome-wide data into research on DT. Findings from GWAS have highlighted the roles of genes involved in both the excitatory and the inhibitory neurotransmission pathways (Liu et al., 2018). Moreover, findings from GWAS-based polygenic score (PS) analysis have provided evidence for the correlations of DT with multiple psychiatric disorders and behavioral traits (Li et al., 2020).

    View all citing articles on Scopus
    1

    These authors contributed equally to the work.

    View full text