Habenula neural circuitry drives negative self-cognitions

Habenula neural circuitry drives negative self-cognitions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Habenula neural circuitry drives negative self-cognitions Po-Han Kung, Matthew Greaves, Eva Guerrero-Hreins, Ben Harrison, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5634827/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 May, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Self-related cognitions are integral to personal identity and psychological wellbeing. Persistent engagement with negative self-cognitions can precipitate mental ill health; whereas the ability to restructure them is protective. Here, we leverage ultra-high field 7T fMRI and dynamic causal modelling to characterise a negative self-cognition network centred on the habenula – a small midbrain region linked to the encoding of punishment and negative outcomes. We model habenula effective connectivity in a discovery sample of healthy young adults ( n = 48) and in a replication cohort ( n = 56) using a novel cognitive restructuring task during which participants repeated or restructured negative self-cognitions. The restructuring of negative self-cognitions elicits an excitatory effect from the habenula to the posterior orbitofrontal cortex that is reliably observed across both samples. Furthermore, we identify an excitatory effect of the habenula on the posterior cingulate cortex during both the repeating and restructuring of self-cognitions. Our study provides the first evidence in humans demonstrating the habenula’s contribution to processing self-cognitions. These findings yield novel insights into habenula’s function beyond processing external reward/punishment to include abstract internal experiences. Biological sciences/Neuroscience/Cognitive neuroscience Biological sciences/Neuroscience/Computational neuroscience/Network models Health sciences/Health care/Medical imaging/Magnetic resonance imaging Biological sciences/Neuroscience/Neural circuits Biological sciences/Psychology/Human behaviour Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Self-cognitions are thoughts and beliefs about how an individual perceives themselves, their attributes, as well as their relationship with others and the world 1 . These are often derived from personal experiences, helping to form a cohesive narrative of “Who am I?” and “What am I like?” that defines the distinctly human phenomenon of having a sense of ‘self’ 2 . Importantly, self-cognitions influence how one evaluates and interprets past events, responds to present circumstances, and predicts future situations 1 . As such, self-cognitions are deeply intertwined with an individual’s affective experience and play a principal role in psychological wellbeing 3 , 4 . For instance, persistent engagement with negative self-cognitions in the form of repetitive negative thinking has been shown to contribute to mental ill health 5 , 6 . In contrast, the ability to restructure and update negative self-cognitions with more adaptive narratives can alleviate negative affect and act as a protective factor for mental wellbeing 7 , 8 . Despite their significance to mental wellbeing, the brain mechanisms supporting the higher-order processing of self-cognitions remain largely unexplored. Understanding the neural mechanisms of negative self-cognitions would provide valuable insight into the biological basis of maladaptive thinking patterns, such as rumination, which contribute to depression and anxiety disorders 6 , 9 , 10 . The habenula – a pair of small midbrain nuclei adjacent to the posterior mediodorsal thalamus – may play a role in encoding negative self-cognitions owing to its distinctive function in the processing of other negative stimuli 11 , 12 . Converging animal and human studies have shown that habenula activity increases in response to the omission of expected reward 13 , 14 and to the delivery of punishment 15 . Conversely, the majority of habenula neurons show reduced firing during unexpected reward receipt and when facing reward-predictive cues 16 , 17 . The habenula’s unique function in signalling negative valence and non-reward events has led to it being recognised as the ‘anti-reward’ centre of the brain 18 . In particular, rodent models have shown that neurochemical activation of the habenula induces depression-like symptoms characterised by reduced mobility and sucrose preference, which can be alleviated via pharmacological inhibition of the habenula 19 . These findings have been complemented by human studies reporting increased habenula volume and activity in depression 20 , 21 , as well as intensified habenula activity in response to negative feedback during cognitive tasks 22 . The habenula’s contribution to shaping affective and behavioural responses to negative stimuli is likely underpinned by its extensive connectivity bridging the forebrain to midbrain monoamine systems 23 , 24 . Specifically, the habenula receives efferent projections from the medial prefrontal cortex (mPFC), the basal ganglia (e.g., globus pallidus), and the lateral hypothalamus, which supply information related to an individual’s motivational state 19 , 25 . The habenula, in turn, modulates downstream neurotransmission to shape cognition and behaviour via bidirectional connections with the ventral tegmental area (VTA), substantia nigra compacta, and the raphe nucleus 26 , 27 , 28 . In rats, synaptic potentiation of habenula neurons projecting to the VTA has been found to modulate learned helplessness 29 , and elevated habenula activity has been shown to induce depressive behaviours by reducing serotonin transmission from the dorsal raphe 30 , 31 . Moreover, signals transmitted from the habenula to the mPFC via VTA dopaminergic neurons mediates conditioned place aversion in rats 32 , 33 . Relatedly, interactions between habenula neurons and the anterior cingulate cortex have been shown to guide choice shifting in response to unrewarding outcomes in primates during a reversal learning task 34 . However, it is unclear if such habenula-mediated functions extend to higher-order cognition, such as negative self-related cognitions. As the processing of negative self-cognitions is a uniquely human process, the involvement of the habenula in encoding and restructuring self-cognitions cannot be tested via animal models. To date, human neuroimaging studies have largely focused on habenula response to primary reward or punishment, such as electric shocks 15 , 27 , 35 , or under task-free/resting-state conditions 36 , 37 , 38 . In addition to reward processing regions, resting-state imaging studies have found that habenula activity is correlated with the activity of the orbitofrontal cortex (OFC), hippocampus, and posterior cingulate cortex (PCC) – key regions in large-scale networks implicated in outcome valuation, memory functioning and self-referential cognition 36 , 38 , 39 . However, a mechanistic account of the habenula’s influence over these regions to support valence attribution and negative self-related cognitions remains undefined. Furthermore, assessing habenula response using standard 3-Tesla functional MRI (fMRI) is challenging as limitations in signal contrast and spatial resolution hinder the accurate delineation of the habenula from nearby structures 26 , 40 . Ultra-high field (7T) MRI can overcome these challenges by providing the superior image resolution and signal-to-noise ratio required to map habenula function 41 , 42 , 43 . In this work, we present the first study characterising the habenula’s involvement in the processing of negative self-cognitions. We sought (1) to characterise habenula activity during the encoding and restructuring of negative self-cognitions and (2) to map the directional influence between the habenula and regions implicated in negative self-cognition processing via dynamic causal modelling (DCM). Under a Bayesian framework, DCM uses a neurobiologically informed generative model to infer the causal excitatory and inhibitory effects brain regions have on one another (i.e., effective connectivity), as well as to classify how these interactions are modulated by experimental tasks 44 , 45 , 46 . Given the habenula’s role in signalling negative valence, we hypothesised that habenula activity would increase during the encoding of negative self-cognitions, and that this would be heightened when participants repeat negative cognitions as opposed to restructuring them. We further hypothesised that both the repeating and restructuring of negative self-cognitions would positively modulate connectivity within our habenula-centric network. In addition, we examined the extent to which habenula connectivity during negative self-cognition processing was associated with participants’ endorsement of negative cognitions, as well as their tendency to engage in repetitive negative thinking. We carried out our analyses using a discovery sample including 48 healthy participants, and tested the replicability of our obtained findings in an independent replication sample comprising 65 healthy participants. Using 7T fMRI, we demonstrate that the repetition of negative self-cognitions elicits heightened activity in the habenula compared to restructuring, alongside regions implicated in self-directed thinking (e.g., PCC), outcome valuation (e.g., OFC), and memory (e.g., hippocampus). DCM analyses in the discovery sample reveal that the habenula exerted an excitatory influence on the PCC during both the restructuring and repeating of negative cognitions. In contrast, restructuring negative self-cognitions is characterised by the habenula having an excitatory modulatory effect on the OFC. Our replication sample corroborates this excitatory effect from the habenula to the OFC during the restructuring of negative self-cognitions, providing novel and consistent evidence for the habenula’s involvement in processing negatively valanced self-cognitions that extend beyond its previously limited role in encoding primary punishment and reward. Results Cognitive restructuring paradigm All participants completed a novel block-design cognitive restructuring paradigm 47 while undergoing fMRI scanning (Fig. 1 ; detailed in Methods). Prior to scanning, participants received training on how to restructure negative self-cognitions using Socratic questioning techniques, such as logical reasoning and perspective shifting 48 . At the start of each block, participants were presented with a commonly reported negative self-cognition statement 49 , 50 and given the option to restructure or to repeat each statement. For half of the task blocks, participants restructured the negative self-cognition statements (challenge condition) using previously taught Socratic questioning techniques 48 . For the other half of the task blocks, the participants silently repeated the statement to themselves 5 without engaging in any conscious attempts to refute the negative self-cognitions (repeat condition). Each task block concluded with a fixation cross (rest condition) before the next block began. Prior to and following scanning, participants rated the extent to which they agreed with the presented negative self-cognition statements. Participants also completed the Perseverative Thinking Questionnaire 51 to assess repetitive negative thinking tendencies (reported in Supplementary Table 1). Habenula activity & effective connectivity during negative self-cognition processing Mass-univariate general linear model (GLM) activation analysis revealed that the habenula had increased activity during the repeating of negative self-cognitions compared to restructuring, in addition to the right PCC, right hippocampus, and the right pOFC (Fig. 2 ; Supplementary Fig. 1 & Table 2). As illustrated in Fig. 2 d, habenula response increased during both the repeating and restructuring of negative self-cognitions relative to rest. Complete GLM activation results are reported in Supplementary Table 2. Based on these initial activation results ( n = 48) and past neuroimaging evidence 36 , 38 , 39 , a DCM network including the habenula, right PCC, right hippocampus, and right pOFC as regions-of-interest (model nodes) was inverted for each participant to infer the modulatory influence of negative self-cognition processing on habenula effective connectivity (Fig. 2 e-f) 52 , 53 . Our hypothesised network structure assumed the presentation of negative self-cognition statements as driving input into all regions-of-interest. To probe the effect of habenula activity on the neuronal response of other network regions, we modelled the habenula’s bidirectional pathways to and from the other network nodes, in addition to their self-connections. For each of these pathways, we estimated their (1) intrinsic connectivity, which represent the context-independent interaction between the network regions, i.e., average effective connectivity across the entire paradigm; and (2) the modulatory effects of repeating or restructuring negative self-cognitions on interregional effective connectivity. Group-level effects were summarised using Parametric Empirical Bayes (PEB) 53 and thresholded at posterior probability > .95. The strength of effectivity connectivity is represented as a partial derivative, measured in hertz (Hz), that quantifies the rate at which neuronal activity in one region changes with respect to neuronal activity in another region (intrinsic connectivity) or due to an experimental input (modulatory effect). DCM inversion ( n = 45) and PEB revealed that both the repeating and restructuring of negative self-cognitions positively modulated the connectivity from the habenula to the PCC, such that the habenula exerted an excitatory effect on the PCC during both task conditions (Fig. 3 a-b). The habenula-to-pOFC pathway was positively modulated by the restructuring of negative self-cognitions, suggesting that the habenula upregulated pOFC activity during cognitive restructuring but not the repeat condition. With regards to intrinsic effective connectivity, the PCC and the pOFC had an excitatory influence on the habenula when averaged across the entire cognitive restructuring task (Fig. 3 a); whereas the habenula had an inhibitory effect on the activity of the PCC. Complete Bayesian model-averaged parameter estimates, including posterior expectation, posterior covariance and posterior probability are reported in Table 1 . Table 1 Bayesian model-averaged DCM parameters for endogenous and modulatory connections in the discovery sample Connection Ep Cp PP Endogenous connections a (A-matrix) Habenula → Habenula -0.57 0.0016 1.00* Habenula → PCC -0.18 0.0006 1.00* Habenula → Hippocampus < 0.01 < 0.0001 .00 Habenula → pOFC < 0.01 < 0.0001 .00 PCC → PCC -0.49 0.0017 1.00* PCC → Habenula 0.07 0.0005 .99* Hippocampus → Hippocampus -0.44 0.0020 1.00* Hippocampus → Habenula -0.09 0.0016 .93 pOFC → pOFC -0.27 0.0021 1.00* pOFC → Habenula 0.13 0.0011 1.00* Modulatory connections b (B-matrix) Challenge (CHAL) Habenula → PCC 0.94 0.0121 1.00* Habenula → Hippocampus < 0.01 < 0.0001 .00 Habenula → pOFC 0.33 0.0052 1.00* PCC → Habenula < 0.01 < 0.0001 .00 Hippocampus → Habenula < -0.01 < 0.0001 .00 pOFC → Habenula < 0.01 < 0.0001 .00 Repeat (REP) Habenula → PCC 0.44 0.0099 1.00* Habenula → Hippocampus < -0.01 < 0.0001 .00 Habenula → pOFC < -0.01 < 0.0001 .00 PCC → Habenula < 0.01 < 0.0001 .00 Hippocampus → Habenula < 0.01 < 0.0001 .00 pOFC → Habenula < 0.01 < 0.0001 .00 a Endogenous parameters reflect the average effective coupling between regions across experimental conditions (context-independent). b Modulatory parameters reflect the changes in effective coupling between regions induced by cognitive reappraisal (content-dependent). *Posterior probability (PP) exceeding .95 provides sufficient evidence for a non-zero group effect 53 . Cp posterior covariance, Ep posterior expectation, pOFC posterior orbitofrontal cortex, PP posterior probability, PCC posterior cingulate cortex. Two additional PEB models including participant’s negative self-cognition endorsement and PTQ total scores as covariates did not show sufficient evidence to suggest that neither negative self-cognition endorsement nor perseverative thinking modulated habenula network dynamics (see Supplementary Table 3). Independent sample replication & randomised 5-fold validation Independent sample replication & randomised 5-fold validation We evaluated the out-of-sample validity of our effective connectivity findings using an independent replication dataset ( n = 56). Brain activation patterns in the replication sample during both the repeating and restructuring of negative self-cognitions were largely consistent with the discovery sample (Supplementary Fig. 2, 3 & Table 4). For the replication sample, the specification of both the DCM and group-level PEB model replicated the procedures used in the discovery model. However, we leveraged the Bayesian model-averaged group-level posterior distribution from the discovery model, specified by its mean and covariance, as an empirical prior distribution over the effective connectivity parameters of the replication PEB model 54 , 55 , 56 . This allowed us to inform the inversion of the replication PEB model and test the discovery model parameters in the independent dataset. The use of informed priors capitalises on the inherent advantage of the empirical Bayesian framework to test the validity of our findings – that is, whether the same effective connectivity architecture is replicated in the independent dataset, given prior knowledge on the network dynamics derived from the discovery sample 53 , 57 . Here, we report the DCM results from this informed replication model and summarise the parameter estimates in Supplementary Table 5. We also include the results of the non-informed model in Supplementary Table 6. The positive modulatory effects of the restructuring of negative self-cognitions on the habenula-to-pOFC pathway were identified again in the replication model, such that the habenula exerted an excitatory influence on the pOFC during cognitive restructuring (Fig. 4 ). The modulatory effects associated with the habenula-to-PCC pathway during restructuring and repeating were positive, though they did not surpass the posterior probability > .95 threshold. To further assess whether these findings were influenced by individual participant variance, we conducted a randomised stratified 5-fold validation using the combined dataset (including both the discovery and replication groups) and examined the consistency of connectivity results across the validation subsamples. As reported in Supplementary Table 7, the subsamples were comparable on sex (χ 2 = 7.60, P bonf.−corrected = .535), age ( F 4,96 = 0.69, P bonf.−corrected = 1.000), perseverative thinking ( F 4,96 = 1.39, P bonf.−corrected = 1.000), and endorsement of negative self-cognitions ( F 4,96 = 0.20, P bonf.−corrected = 1.000). Consistent with the replication model described above, the 5-fold validation group-level DCM models were furnished with the empirical prior distribution derived from the posterior distribution of the discovery model. Through this procedure, we found very strong evidence (posterior probability > .95) supporting the positive modulatory connectivity of the habenula-to-pOFC during the restructuring of negative self-cognitions in 4 out of the 5 subsamples (Supplementary Fig. 4). Additionally, very strong evidence (posterior probability > .95) of positive modulation of the habenula-to-PCC connection during the restructuring and repeating conditions were replicated in 3 out of the 5 subsamples, upholding the reliability of our connectivity results despite moderate levels of inter-sample variability. Discussion In this study, we combined DCM and 7T fMRI to infer the role of a habenula-centric circuitry during the processing of negative self-cognitions. In line with our hypothesis, we observed increased habenula activity during the repeating of negative self-cognitions compared to restructuring. Using data from two independently acquired samples, as well as a randomised 5-fold validation, we identified a reliable network structure revealing excitatory effective connectivity from the habenula to the pOFC during the restructuring of negative self-cognitions. In the discovery sample, we identified excitatory effective connectivity from the habenula to the PCC during both the repeating and restructuring of negative self-cognitions. These findings provide novel insights into the habenula’s functional influence on key nodes of the default mode network and cognitive control network to support self-related higher-order cognitions in humans, thereby broadening our understanding of habenula function to encompass domains not limited to external primary reward or punishment. Although both the restructure and repeat task conditions modulated habenula activity, it is worth noting that habenula response was heightened during the repeating of negative self-cognitions. As negative self-cognitions were not consciously restructured and reduced in intensity during the repeat condition, increased habenula activity may reflect the sustained signalling of negative valence induced by the repetition of negative self-cognition statements. This aligns with the understanding that habenula activity encodes the negative motivational value of external punishment or unrewarding outcomes to influence behavioural response 18 , 58 , suggesting parallel neural processes between the encoding of the aversiveness of negative self-cognitions and negative reward signalling. This valence-based information may then be integrated into higher-order cognition subserved by other cortical systems 59 . Our model demonstrated that both the restructuring and repeating of negative self-cognitions positively modulated habenula-to-PCC connectivity, such that the habenula consistently exerted an excitatory influence on the PCC during engagement with negative self-cognitions. As a core node of the default mode network, the PCC has been posited to play a coordinating role in the flexible attentional switch between internal and external environments 60 , 61 . Through interactions with the frontoparietal executive control regions and the salience network (e.g., anterior cingulate cortex, anterior insula), the PCC receives information regarding the personal relevance of the task at-hand to inform attentional resource allocation 62 , 63 , with sustained PCC activity facilitating more self-oriented cognition 64 . Crosstalk between the habenula and the PCC may similarly allow the incorporation of valence and motivational value to tilt attentional balance towards self-referential processes when faced with negative self-related cognitions. This hypothesis accords with the view that connectivity between the habenula and the default mode network potentially reflects an integrative self-monitoring process, with the value of negative stimuli being signalled by the habenula 39 . Moreover, the PCC is integral to the generation of a unitary representation of the self and performs a gating function by which self-conceptualisation enters conscious awareness 60 , 65 . This is particularly relevant during autobiographical memory recall where the PCC-mediated self-conceptualisation represents egocentric information associated with prior experiences 62 , 64 . Via the habenula-to-PCC pathway, negative valence encoded in the habenula may be attributed to mental representations of the self during associations triggered by the internal recital of negative statements or during the conscious recollection of personal experiences required to refute them. The habenula exerted a distinct excitatory effect on the pOFC during the restructuring of negative self-cognitions – an effect that was reliably detected in both the discovery and replication samples. Restructuring negative self-cognitions not only requires the sustaining of complex self-concepts, subserved by the default mode network 65 , but also the manipulation of self-representations, which has been shown to involve frontostriatal valuation and cognitive control circuits 47 , 66 . The observed excitatory connectivity from the habenula to the pOFC concurs with these reports and suggests a previously undescribed role of the habenula in shaping adaptive responses to negative cognitions. Besides negative valence, the habenula is known to be sensitive to trial-to-trial feedback for reward outcomes 67 , 68 , and to contribute to the flexible modification of action strategies in reaction to aversive outcomes 14 , 69 . These findings point to the habenula’s contribution to tracking the effectiveness of behavioural responses in relation to outcome expectations and feedback throughout changing contexts. In this regard, excitatory connectivity from the habenula to the pOFC may act as a pathway through which the expectation and outcome of the restructuring effort is transmitted from the midbrain to the prefrontal cortex. Our findings are in accordance with the growing consensus that the OFC constructs and maintains a cognitive map defining the current task space, in which multiple sources of information relevant to decision-making (e.g., action-outcome value, emotion, memory) are synthesised 70 , 71 , 72 . The recruitment of the OFC is necessary in situations requiring mental simulation or future inferences, where values and predictions of possible outcomes associated with each choice options need to be computed with continuously updated information 73 , 74 . During the cognitive restructuring paradigm used in this study, participants were presented with a different statement on each trial and could not rely on previously formulated arguments to restructure the negative self-cognitions. Rather, participants need to dynamically adjust their cognitive strategies and to conceive new rebuttals in response to different self-cognition statements. Each restructuring strategy may represent an alternative task state that necessitates OFC-mediated representation and outcome-value computation 72 . This real-time evaluation likely incorporates the moment-by-moment feedback about the expected and actual effectiveness of the restructuring strategy encoded by the habenula 69 , 75 , a process potentially subserved by the restructuring-induced positive modulation of the habenula-to-pOFC excitatory connectivity. Mizumori and Baker 76 recently synthesised findings from animal models to hypothesise that the habenula integrates action-outcome valuation from the mPFC and information about the organism’s internal state from the subcortex (e.g., lateral hypothalamus, entopeduncular nucleus) to signal the effectiveness of behaviours in relation to a contextual goal 76 , 77 . They proposed that the habenula encodes the decision to continue or alter the current course of action while simultaneously relaying this information to the hippocampus and the mPFC where subsequent actions may be updated and evaluated 76 . How this model applies to humans remains unknown. Nevertheless, our current findings support the role of habenula-frontal cortex interactions in the adaptive processing of negative cognitions and significantly expand the evidence base for the human habenula's role in complex, higher-order cognitive processes 78 , 79 . Contrary to our hypotheses, we could not find sufficient evidence to support modulations in connectivity between the habenula and the hippocampus during the processing of negative self-cognitions. Insights into habenula-hippocampus interaction during mnemonic processing has primarily been derived from animal models of conditioned fear 80 , spatial and working memory 77 , 81 , 82 . How the habenula may support memory processes in humans, such as episodic memory recall for the restructuring of negative self-cognitions based on personal history, and working memory to sustain this process, remains an open question to be explored using different paradigms. We note some limitations of this work. First, the current samples included healthy participants who tended to report low levels of negative self-cognition endorsement and perseverative thinking. Thus, a floor effect may have impacted our ability to detect meaningful relationships between individual endorsement of negative self-cognitions and habenula connectivity. Relatedly, the lack of model evidence for an association between the perseverative thinking and negative self-cognition measures and habenula connectivity precluded inferences on the behavioural implication of habenula connectivity variations in healthy individuals. While the present study focused on establishing a normative role of the habenula in negative self-cognition processing, future studies would benefit from extending this to populations with elevated levels of maladaptive cognitions (e.g., people experiencing psychopathology) 6 . Lastly, there were minor methodological differences between our discovery and replication datasets that may have contributed to variance in our connectivity results. Despite these dissimilarities, we observed consistent modulation of the habenula connectivity during the restructuring of negative self-cognitions across the samples, which upholds the reliability of these findings. Here, we present the first evidence in humans demonstrating habenula involvement in the higher-order processing of negative self-cognitions. We showed that habenula activity was modulated by the repeating and restructuring of negative self-cognitions. Using DCM in two independent samples, our model elucidated the directional interplay between the habenula, PCC, and the OFC, which prospectively underpins negative self-conceptualisation and the value-guided restructuring of negative self-cognitions. These findings advance our current understanding of the habenula’s role in negative stimuli processing beyond primary reward and punishment to include abstract internal experiences. A mechanistic account of habenula functioning lays the foundation for future work examining neural vulnerabilities contributing to maladaptive thinking patterns and whether the habenula represents a treatment target to alleviate entrenched negative self-cognitions that do not respond to conventional psychotherapy alone. Methods Participants For our discovery sample, we recruited 57 healthy adults from the community via online advertisements. Inclusion criteria included: 1) being between the age of 18 and 40 years; 2) having no MRI contraindications (e.g., pregnancy, metallic implants or claustrophobia); 3) willingness to comply with the scanning centre’s healthy and safety policies (e.g., received full course of SARS-CoV-2 vaccination); 4) fluency in written and verbal English; 5) being capable of complying with study instructions. Participants were excluded if they 1) had a diagnosis of any mental disorder at the time of study participation; (2) have a past history of eating disorders, psychotic disorders, obsessive-compulsive disorder, or bipolar disorders based on The Mini International Neuropsychiatric Interview (MINI, English version 7.0.2) 83 for the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) 84 ; 3) have been diagnosed with autism spectrum disorder; 4) have major hearing or sight difficulties; or 5) have a medical or neurological condition for which they are on medication. For our independent replication sample, we obtained MRI data from 83 healthy adults, which have been described in previous reports 47 , 66 , 85 . In brief, participants were eligible if they 1) aged between 18 and 40 years; 2) did not meet criteria for any mental disorders as screened using the MINI; 3) had no MRI contraindications; and 4) were proficient in English and had normal or corrected-to-normal vision. All participants provided written informed consent and attended one testing session at the Melbourne Brain Centre Imaging Unit (The University of Melbourne, Parkville, Victoria, Australia). This study was approved by the University of Melbourne Human Research Ethics Committee (HREC 22347, 2056265). Nine and 18 participants were initially excluded from the discovery and replication samples, respectively, due to: technical errors during MRI acquisition (discovery: 1, replication: 3), participant not completing or incorrectly completing the fMRI paradigm (discovery: 7, replication: 5), and excessive head motion (discovery: 1, replication: 10). Thus, 48 participants from the discovery and 65 participants from the replication samples were included in the GLM activation analysis. An additional 3 participants from the discovery and 9 participants from the replication samples were excluded from the DCM analysis due to a failure to extract a valid time series from the regions-of-interest, resulting in 45 discovery group participants and 56 replication group participants being included in the final DCM analysis. Self-report measures Demographic information. Age, sex (i.e., assigned sex at-birth), and ethno-cultural group based on participant self-report are summarised in Supplementary Table 1. Perseverative thinking questionnaire (PTQ) 51 . The PTQ is a 15-item questionnaire assessing individuals’ general propensity to engage in repetitive negative thinking and its impact. Each item is rated on a 4-point Likert scale (0 = never; 4 = almost always), with higher total scores indicating stronger perseverative thinking tendencies. The PTQ was administered to all eligible participants prior to the MRI scanning session. Challenging negative beliefs task questionnaire (CNBTQ). The CNBTQ forms part of the cognitive restructuring fMRI task (see section below for detailed task description) and was administered prior to scanning and following scanning. The CNBTQ has participants rate the extent to which they endorse the negative self-cognition statements presented during the cognitive restructuring paradigm on a 7-point Likert scale (1 = strongly disagree; 7 = strongly agree). The difference in pre- and post-task ratings was used to index task-invoked shifts in negative self-cognition endorsement. Participants in the replication sample were only asked to provide post-task ratings on the statements that were restructured during the task, whereas participants in the discovery sample provided post-task ratings to all statements. Behavioural analyses Demographic and behavioural variables were analysed using SPSS v.27 (IBMCorp., Armonk, NY). Changes in negative self-cognition endorsement were compared within-group using single sample t-tests (2-tailed), and between-groups using independent sample t-tests (2-tailed). Likewise, two-tailed independent sample t-tests were used to assess differences in demographics and behavioural variables across the samples, whereas two-tailed Mann-Whitney U tests were used when there was evidence for heteroscedasticity. One-way analysis of variance (ANOVA) was used when three or more subgroups were compared. Two-tailed Pearson's chi-squared tests were adopted to evaluate categorical variables between the groups. Bonferroni correction was applied to account for multiple comparisons. Cognitive restructuring paradigm As described in Steward et al. 47 , prior to scanning, participants received training to restructure negative cognitions using Socratic questioning techniques that mimic those practised in cognitive psychotherapy, such as recalling personal experiences that refute negatively biased self-conceptions 49 . Once research staff verified that the participant could correctly carry out Socratic questioning with the goal to restructure negative self-cognitions, participants completed the paradigm during MRI scanning. The cognitive restructuring task comprised one run of 24 blocks (Fig. 1 ). In each block, the participants were first shown a statement on screen of common negative self-cognitions about the self, food, and body image reported in the cognitive behavioural therapy and psychopathology literature (e.g., “I am incompetent in the things I do”, “My value depends on my body shape”; Supplementary Table 8) 50 , 86 . Next, participants decided via a button press whether they would repeat or restructure the presented statement using pre-trained strategies. Participants were instructed to restructure and repeat an equal number of statements. After 9 seconds, the participants were shown the same statement and instructed to engage in their chosen strategy for 12 seconds (restructure = challenge condition; repeat = repeat condition). A fixation cross was then presented for 6 seconds (rest condition) before the next block began with a new statement. The cognitive restructuring paradigm used in the discovery and replication samples were identical, except that the replication group completed an abbreviated version containing one run of 16 blocks that did not include negative self-cognition statements about food and body image. fMRI image acquisition Imaging for the discovery and replication datasets was conducted on a 7-Tesla research scanner (Siemens Healthcare, Erlangen, Germany) equipped with an 8Tx/32Rx and 1Tx/32Rx head coil, respectively (Nova Medical Inc., Wilmington, MA, USA). The functional sequence was consistent across the two samples and consisted of a multi-band (factor = 6) and GRAPPA (R = 2) accelerated GE-EPI sequence 87 in the steady state (TR = 800 ms; TE = 22.2 ms; pulse angle = 45°; field of view = 20.8 cm; acquisition matrix = 130 × 130-pixel; slice thickness = 1.6 mm, no gap). Eighty-four interleaved axial slices were acquired along the anterior-posterior commissure line. In total, 946 and 628 whole-brain EPI volumes were acquired in a single run for the discovery and replication datasets, respectively, corresponding to approximately 12.6 and 8.4 minutes for the two task versions. High-resolution T1-weighted anatomical images were acquired for functional time-series co-registration and individualised habenula segmentation (detailed in section Habenula segmentation & region-of-interest validation ) using the first echo of a multi-echo Magnetization Prepared 2 Rapid Acquisition Gradient Echoes sequence (ME-MP2RAGE) 88 for the discovery dataset (224 interleaved axial slices; TR = 4500 ms; TE = 2.21/4.21/6.15/8.14 ms; inversion time = 700/2700 ms; flip angle = 6/7°; field of view = 24 cm; acquisition matrix = 320 × 320-pixel; slice thickness = 0.75 mm, no gap), and a single-echo MP2RAGE 89 sequence for the replication dataset (224 interleaved sagittal slices; TR = 5000 ms; TE = 2.04 ms; inversion time = 700/2700 ms; flip angle = 4/5°; field of view = 24 cm; acquisition matrix = 320 × 320-pixel; slice thickness = 0.75 mm, no gap). Standard foam pads were used for all participants to minimize head movement during scanning. Respiratory and cardiac recordings were sampled at 50 Hz using a respiratory belt and pulse-oximeter. Information derived from these recordings were used for physiological noise correction during image pre-processing. fMRI image pre-processing Imaging data was pre-processed with Statistical Parametric Mapping 12 (SPM12, v7771; Wellcome Trust Centre for Neuroimaging, London) within the MATLAB 2023a environment (The MathWorks Inc., Natick, MA). Each participant’s functional time series was realigned to the mean image to correct for movement during the scan, and all images were resampled using 4th Degree B-Spline interpolation. Individual head motion was assessed with motion fingerprint 90 . Participants were excluded if they had a mean total displacement over 1.6 mm (one isotropic voxel size) and/or a maximum scan-to-scan displacement exceeding 2 mm. We imposed a stringent censoring criterion to minimise the effect of motion on signal distortion from the habenula. Each participant’s anatomical T1 image was co-registered to their mean functional image, segmented and normalized to the International Consortium for Brain Mapping (ICBM) European brain template using the unified segmentation approach plus Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) 91 . Lastly, the functional images were spatially normalised with the DARTEL flow fields and smoothed with a 2 mm full width at half maximum (FWHM) Gaussian kernel to preserve spatial specificity. Cardiac and respiratory recordings were modelled using the PhysIO toolbox 92 to account for physiological noise 93 . Specifically, the Retrospective Image-based Correction function (RETROICOR) 94 , respiratory response function (RRF) 95 and cardiac response function (CRF) 96 were incorporated to correct for both periodic and variable effects of heartbeat and breathing, as well as their interaction, on BOLD signal. Anatomical component correction (aCompCor) 97 was applied to account for non-neural signal, whereby the mean and first principal components of the time series originating from the white matter (WM) and cerebrospinal fluid (CSF) were extracted using individualised DARTEL tissue maps. Habenula segmentation & region-of-interest (ROI) validation Individual-specific habenula mask was generated in the native space using each participant’s high-resolution anatomical image (0.75 mm isotropic) via a validated, fully automated segmentation algorithm (Multiple Automatically Generated Templates Brain Segmentation Algorithm; MAGeTbrain) 98 , 99 . A schematic of our image preparation pipeline is included in the Supplementary Fig. 5. Using participant anatomical images and a set of habenula atlases as input, MAGeTbrain first propagates the segmentation of the atlas images to each subject image via a multi-stage image registration procedure to yield a large number of candidate segmentations for each individual (5 atlases \(\:\times\:\:\) 21 templates = 105). Next, the candidate segmentations are fused via majority vote at each voxel, whereby the most frequently occurring label (habenula vs. non-habenula) was adopted in the final output segmentation. This approach reduces potential biases related to atlas or rater inconsistency, allows for neuroanatomical variability 40 , and has been shown to produce reliable habenula volume estimates across diverse populations and image acquisition parameters 99 . The MAGeTbrain algorithm produced a binary anatomical-resolution habenula mask in the native space, which we then used to generate functional-resolution habenula ROIs in the standard space for fMRI analyses. We employed an iterative volume optimisation strategy adapted from Ely et al. 36 , 39 to minimise the impact of interpolation during image down-sampling, as well as the increased susceptibility to non-neural noise at a reduced resolution, to produce the most feasibly precise representation of each individual’s habenula ROI in the functional space (Fig. 5 ): 1) Each anatomical habenula mask produced by MAGeTbrain, as well as their corresponding DARTEL segmented CSF mask, were resampled with trilinear interpolation to the functional resolution (2 mm isotropic) based on the mean functional image (Fig. 5 a), normalised to the standard space using their corresponding DARTEL flowfields, and binarised at a conventional threshold of 0.2 (Fig. 5 c), which resulted in habenula and CSF masks showing good spatial consistency with the original anatomical masks. 2) To maximally reduce signal contamination from the CSF, voxels in the resampled habenula mask that overlap with the resampled CSF mask (binarising threshold = 0.2) were removed. 3) The volume of the adjusted habenula mask from Step 2 was compared to the reference habenula segmentation image (i.e., subject-specific normalised habenula mask in anatomical resolution; Fig. 5 b). 4a) If the volume of the adjusted habenula mask from Step 2 fell within ± 10% of the reference image, it was accepted for use as an ROI in subsequent analyses. The volume criterion was designed to optimally approximate our functional habenula ROI to the high-resolution segmentation, while taking into consideration individual anatomical variability. 4b) If the volume was above ± 10% that of the reference image, it was rejected and Step 1 was repeated with the binarising threshold adjusted upwards or downwards by 0.01, producing a slightly smaller or larger habenula mask. Steps 2–3 were then repeated. This process was iterated until the habenula mask at functional resolution was accepted (Step 4a). This iterative volume optimisation generated individualised habenula masks with excellent spatial specificity confirmed via visual inspection and volume estimates comparable with previous reports (i.e., 30–60 mm 3 combined; discovery sample: Mean = 60.93 mm 3 , 95% CI = [57.72, 64.13]; replication sample: Mean = 48.40 mm 3 , 95% CI = [46.35, 50.44]) 26 , 99 , 100 , 101 . We further validated the functional specificity of our subject-specific habenula ROI via a resting-state seed-to-whole brain functional connectivity analysis conducted in CONN22a 102 . Please see Supplementary Methods for detailed information on image acquisition, pre-processing, and analysis. Using the individualised habenula ROI as a seed, we replicated the connectivity patterns reported in past studies at standard and ultra-high field strengths (Fig. 5 d) 36 , 38 , 39 , highlighting habenula connectivity with structures such as the ventral tegmental area, thalamus, insula, anterior through to the posterior cingulate cortex, and the supplementary motor area. General Linear Modelling (GLM) analysis Single-subject (first-level) contrast images were estimated for Challenge > Repeat and Repeat > Challenge (i.e., the entire 12s when participants either restructured or repeated the negative self-cognition statements) to characterise changes in brain activation associated with the restructuring and repeating of negative self-cognitions, respectively. Each participant’s pre-processed timeseries, 6 realignment parameters for motion artifacts, and 24 physiological noise regressors (6 cardiac, 8 respiratory, 4 cardiac \(\:\times\:\:\) respiratory regressors, 1 RRF, 1 CRF, top principal components and mean time series of the CSF and WM) were included in the GLM analysis, with the onset times for each condition event specified and convolved with the SPM canonical hemodynamic response function (HRF). Low-frequency fluctuation was high-pass filtered at 128Hz. The FAST method was used to estimate temporal autocorrelation resulting from our sub-second TR 103 . The first-level contrast images were brought forward to a group-level GLM (one-sample t-test, one-tailed). All GLM analyses were thresholded at a whole brain, false discovery rate (FDR) corrected P FDR < 0.05, K E ≥ 10 voxels. Dynamic Causal Modelling (DCM) DCM is a Bayesian framework used to infer the directional and causal influence that brain regions exert on one another (i.e., effective connectivity) 45 . DCM enables inferences to be made by simulating neuroimaging timeseries via a generative model that is grounded in empirical knowledge of neuronal processes and how they translate to observable responses (e.g., BOLD signal) 44 , 104 . Within a set of researcher-specified hypotheses regarding network structure, DCM estimates connectivity parameters through model inversion in a manner that optimises the trade-off between model fit to the observed data and model complexity 52 . This balance is reflected by free energy, which approximates the (log) model evidence used for model comparison and hypothesis testing between competing models 105 . The model with the most positive free energy reflects a parsimonious and physiologically plausible account of the dynamic interactions between unobserved neuronal populations 106 . Model space & timeseries extraction. Our DCM network included the bilateral habenula as a single ROI, as well as brain regions that were co-activated during Repeat > Challenge that have structural/functional connectivity with the habenula based on prior literature 24 , 36 , 38 . Specifically, the right PCC, hippocampus, and pOFC were included due to their central role in self-referential thinking, episodic memory and learning, as well as adaptive processes that are engaged during negative self-cognition processing. Representative time series (volume-of-interest) were extracted from these areas for each of the subjects following published guidelines 52 . The habenula ROI was delineated with the individualised mask described above. Whereas all other network region ROIs were defined as a 4 mm radius sphere centred around the individual neural activation maxima under the Repeat > Challenge contrast, constrained to be within 8 mm from each dataset’s group peak (Supplementary Table 9). Volume-of-interest from each of these regions were calculated using SPM as the principal eigenvariate of all voxels within the respective ROI that showed meaningful activation for the task contrast ( p < .05, uncorrected) at the single subject level. In the case when an ROI contains no voxel surpassing the preset threshold, the statistical threshold incrementally relaxed to p < .5 (uncorrected) until a peak coordinate can be identified. This approach ensured minimal exclusion of participant data from our analyses as subjects lacking strong responses in a brain region or experimental condition may nevertheless provide useful information about other regions, conditions, and individual variability 52 . Of note, the contrast image used for VOI extraction was denoised with the motion and physiological nuisance regressors described in the General Linear Model section above, except the aCompCor components. This adjustment took into consideration the explicit removal of the CSF voxels from our individualised habenula mask and habenula’s high WM density to prevent overcorrection of the extracted timeseries, while minimising non-neural noise. Individual timeseries were additionally pre-whitened to mitigate serial correlations, high-pass filtered, and nuisance effects not covered by the Effects of Interest F-contrast are regressed out of the timeseries (i.e. ‘adjusted’ to the F-contrast). Using this procedure, we extracted a complete set of VOIs for 45 and 56 individuals from the discovery and replication cohorts, respectively, which were then included in the connectivity analyses. Model specification & estimation. Our model was specified with SPM DCM 12.5 and featured 1) the endogenous connections between and within each target region (A-matrix), 2) the modulatory effect of the task conditions on inter-region connectivity (B-matrix), as well as 3) the driving influence of the task stimuli (C-matrix). All experimental input was mean centred to aid parameter interpretability. As illustrated in Fig. 2 f, the full model assumed bidirectional endogenous connections of the habenula to and from the other regions in addition to their self-inhibitory connectivity, modelling the average connectivity parameters throughout the experiment that is independent of condition effects. We allowed the challenge and repeat conditions to modulate the connectivity between the habenula and other network nodes to characterise changes to habenula connectivity during negative self-cognition processing. Lastly, the aggregate of the 4s of statement presentation and 12s of restructuring/repeating negative self-cognitions was specified as the driving input to all nodes, reflecting the engagement of these regions during the exposure to negative self-cognition statements. This full model was estimated and evaluated for each subject, yielding individual posterior connectivity parameter estimates and their posterior probability. Parametric empirical Bayes. Next, a group-level summary of the connectivity parameters was obtained via Parametric empirical Bayes (PEB) 53 . PEB is a hierarchical model that incorporates both the subject-level parameter estimates and their uncertainty (i.e., posterior covariance) to the group level. Contrary to the standard summary statistics approach, the PEB framework effectively downweights data with noise and uncertain individual estimates to produce more reliable population connectivity estimates 53 , 57 . Additionally, each level of the PEB hierarchy serves as a prior on the estimates of the level below it. This can improve the precision of individual parameter estimates by incorporating knowledge around task effects garnered from the cohort. Our PEB model was designed to investigate the between-subject commonalities in connectivity parameters for each of the samples. The design matrix included an intercept term (single column of ones) denoting the overall mean connectivity 53 . Parameters from both A- and B-matrices were summarised in this model to account for potential conditional dependency. Generic prior distribution were adopted for the discovery model with no assumptions around the strength and variance of network connectivity 52 . Once this full PEB model was inverted, Bayesian Model Reduction (BMR) was used to search and compare the relative evidence of possible reduced models, iteratively pruning parameters that do not contribute to an increase in model evidence 107 , 108 . Bayesian Model Averaging (BMA) was then used to aggregate the parameters of the reduced models, weighted by the corresponding model’s posterior probability, to provide the final group summary 53 . The BMA parameters were thresholded at posterior probability > .95, indicating sufficient evidence for a non-zero group effect. The two PEB models investigating the effect of negative self-cognition endorsement and repetitive negative thinking tendencies on habenula connectivity included the individual pre-task CNBTQ and PTQ total scores, respectively, as covariate regressors in addition to the intercept term. As all regressors were mean-centred, the between-subject effects could be quantified as the addition to or subtraction from the overall mean connectivity estimates 53 . For the replication and 5-fold validation models, the generic prior distribution was substituted in the PEB model with the posterior distribution derived from the discovery model BMA summary. The discovery model posteriors served as 3rd -level empirical priors to constrain the 2nd -level (group-level) estimates of the replication and 5-fold validation models. As the result of BMR applied during the discovery model estimation, only connectivity with a non-zero posterior probability were evaluated in the replication and validation PEB models. This approach allows the testing of the discovery model parameters on the independent dataset and effectively incorporates empirically derived beliefs around network dynamics and their degree of uncertainty to refine the estimation of connectivity parameters in new populations 53 , 57 . Declarations Data availability Deidentified effective connectivity data for this study are publicly available at https://github.com/pohankung/NegativeBeliefs_Habenula_DCM . Source data for Fig. 3 b and 4 b are provided with the submission. Code availability Scripts used to generate the main results and figures of this study are available at https://github.com/pohankung/NegativeBeliefs_Habenula_DCM . Custom code was written in MATLAB 2023a. Statistical Parametric Mapping 12 (SPM12) and FMRIB Software Library (FSL) 6.0.6.5 were used for MRI processing. Competing interests The authors declare no competing interests. Author contributions P.-H.K., B.J.H., and T.S. conceived the general concept of this study, designed the experiment, and developed the model with input from M.D.G. and E.G.-H. P.-H.K., E.G.-H., B.J.H., K.L.F., H.C., P.S., R.M.B., B.A.M., R.K.G., and T.S. aided with data collection and the crafting of the imaging protocol. P.-H.K. and T.S. conducted data analysis with support from M.D.G. B.J.H., C.G.D., K.L.F., P.S., R.M.B., and T.S. provided supervision throughout the study. P.-H.K. and T.S. wrote the original draft of the manuscript. All authors reviewed and approved the final edit of this manuscript. Acknowledgements We thank James Agathos, Carly Beveridge, Lieselotte Claes, Yingliang Dai, Elizabeth Haris, Sevil Ince, Amy Nielson, Mia O’Shea, Tudor Sava, Braden Thai, and Andong Zhou for their contribution to data collection. We acknowledge the technical and scientific assistance of the Australian National Imaging Facility – a National Collaborative Research Infrastructure Strategy (NCRIS) capability at the Melbourne Brain Centre Imaging Unit (MBCIU), The University of Melbourne. The multiband fMRI sequence was generously supported by a research collaboration agreement with CMRR, The University of Minnesota. Siemens Healthineers (Germany) provided the MP2RAGE sequence. This study was supported by the National Health and Medical Research Council of Australia (NHMRC)/Medical Research Future Fund (MRFF) Investigator Grant (MRF1193736), a Brain & Behaviour Research Foundation (BBRF) Young Investigator Grant and a University of Melbourne McKenzie Fellowship to T.S. Collection of the replication dataset was supported by a NHMRC Project Grant (1161897) to B.J.H. and an NHMRC Program Grant (1073041) to K.L.F. References Leary MR, Tangney JP (2012) The self as an organizing construct in the behavioral and social sciences. In: Handbook of self and identity (ed^(eds Leary MR, Tangney JP). 2nd edn. The Guilford Press Gallagher S (2000) Philosophical conceptions of the self: implications for cognitive science. Trends Cogn Sci 4:14–21 Segal ZV (1988) Appraisal of the self-schema construct in cognitive models of depression. Psychol Bull 103:147–162 Hoven M, Luigjes J, Denys D, Rouault M, van Holst RJ (2023) How do confidence and self-beliefs relate in psychopathology: A transdiagnostic approach. Nat Mental Health 1:337–345 Ehring T, Watkins ER (2008) Repetitive negative thinking as a transdiagnostic process. Int J Cogn Ther 1:192–205 Wahl K et al (2019) Is repetitive negative thinking a transdiagnostic process? A comparison of key processes of RNT in depression, generalized anxiety disorder, obsessive-compulsive disorder, and community controls. J Behav Ther Exp Psychiatry 64:45–53 Braun JD, Strunk DR, Sasso KE, Cooper AA (2015) Therapist use of Socratic questioning predicts session-to-session symptom change in cognitive therapy for depression. Behav Res Ther 70:32–37 Ezawa ID, Hollon SD (2023) Cognitive restructuring and psychotherapy outcome: A meta-analytic review. Psychotherapy 60:396–406 McEvoy PM, Watson H, Watkins ER, Nathan P (2013) The relationship between worry, rumination, and comorbidity: Evidence for repetitive negative thinking as a transdiagnostic construct. J Affect Disord 151:313–320 Watkins ER, Roberts H (2020) Reflecting on rumination: Consequences, causes, mechanisms and treatment of rumination. Behav Res Ther 127:103573 Mondoloni S, Mameli M, Congiu M (2022) Reward and aversion encoding in the lateral habenula for innate and learned behaviours. Transl Psychiatry 12:3 Hikosaka O (2010) The habenula: From stress evasion to value-based decision-making. Nat Rev Neurosci 11:503–513 Shabel SJ, Wang C, Monk B, Aronson S, Malinow R (2019) Stress transforms lateral habenula reward responses into punishment signals. Proc. Natl. Acad. Sci. U. S. A. 116, 12488 Sosa R, Mata-Luévanos J, Buenrostro-Jáuregui M (2021) The role of the lateral habenula in inhibitory learning from reward omission. eneuro 8, ENEURO.0016-0021.2021 Lawson RP et al (2014) The habenula encodes negative motivational value associated with primary punishment in humans. Proc. Natl. Acad. Sci. U. S. A. 111, 11858–11863 Matsumoto M, Hikosaka O (2007) Lateral habenula as a source of negative reward signals in dopamine neurons. Nature 447:1111–1115 Weidacker K, Kim S-G, Nord CL, Rua C, Rodgers CT, Voon V (2021) Avoiding monetary loss: A human habenula functional MRI ultra-high field study. Cortex 142:62–73 Ables JL, Park K, Ibañez–Tallon I (2023) Understanding the habenula: A major node in circuits regulating emotion and motivation. Pharmacol Res 190:106734 Boulos L-J, Darcq E, Kieffer BL (2017) Translating the Habenula - From Rodents to Humans. Biol Psychiatry 81:296–305 Schmidt FM et al (2017) Habenula volume increases with disease severity in unmedicated major depressive disorder as revealed by 7T MRI. Eur Arch Psychiatry Clin Neurosci 267:107–115 Proulx CD, Hikosaka O, Malinow R (2014) Reward processing by the lateral habenula in normal and depressive behaviors. Nat Neurosci 17:1146–1152 Lecourtier L, Kelly PH (2007) A conductor hidden in the orchestra? Role of the habenular complex in monoamine transmission and cognition. Neurosci Biobehav Rev 31:658–672 Fakhoury M (2017) The habenula in psychiatric disorders: More than three decades of translational investigation. Neurosci Biobehav Rev 83:721–735 Hu H, Cui Y, Yang Y (2020) Circuits and functions of the lateral habenula in health and in disease. Nat Rev Neurosci 21:277–295 Stamatakis AM, Van Swieten M, Basiri ML, Blair GA, Kantak P, Stuber GD (2016) Lateral hypothalamic area glutamatergic neurons and their projections to the lateral habenula regulate feeding and reward. J Neurosci 36:302 Lawson RP, Drevets WC, Roiser JP (2013) Defining the habenula in human neuroimaging studies. NeuroImage 64:722–727 Hennigan K, D'Ardenne K, McClure SM (2015) Distinct midbrain and habenula pathways are involved in processing aversive events in humans. J Neurosci 35:198–208 Stamatakis Alice M et al (2013) A unique population of ventral tegmental area neurons inhibits the lateral habenula to promote reward. Neuron 80:1039–1053 Li B et al (2011) Synaptic potentiation onto habenula neurons in the learned helplessness model of depression. Nature 470:535–539 Metzger M, Bueno D, Lima LB (2017) The lateral habenula and the serotonergic system. Pharmacol Biochem Behav 162:22–28 Zhao H, Zhang B-L, Yang S-J, Rusak B (2015) The role of lateral habenula–dorsal raphe nucleus circuits in higher brain functions and psychiatric illness. Behav Brain Res 277:89–98 Baker PM et al (2016) The lateral habenula circuitry: Reward processing and cognitive control. J Neurosci 36:11482–11488 Lammel S et al (2012) Input-specific control of reward and aversion in the ventral tegmental area. Nature 491:212–217 Kawai T, Yamada H, Sato N, Takada M, Matsumoto M (2015) Roles of the lateral habenula and anterior cingulate cortex in negative outcome monitoring and behavioral adjustment in nonhuman primates. Neuron 88:792–804 Lawson RP et al (2017) Disrupted habenula function in major depression. Mol Psychiatry 22:202–208 Ely BA, Stern ER, Kim J-w, Gabbay V, Xu J (2019) Detailed mapping of human habenula resting-state functional connectivity. NeuroImage 200:621–634 Zhu Z, Wang S, Lee TMC, Zhang R (2023) Habenula functional connectivity variability increases with disease severity in individuals with major depression. J Affect Disord 333:216–224 Torrisi S, Nord CL, Balderston NL, Roiser JP, Grillon C, Ernst M (2017) Resting state connectivity of the human habenula at ultra-high field. NeuroImage 147:872–879 Ely BA, Xu J, Goodman WK, Lapidus KA, Gabbay V, Stern ER (2016) Resting-state functional connectivity of the human habenula in healthy individuals: Associations with subclinical depression. Hum Brain Mapp 37:2369–2384 Bian B et al (2023) Recent advances in habenula imaging technology: A comprehensive review. J Magn Reson Imaging Dumoulin SO, Fracasso A, van der Zwaag W, Siero JCW, Petridou N (2018) Ultra-high field MRI: Advancing systems neuroscience towards mesoscopic human brain function. NeuroImage 168:345–357 Harrison OK, Guell X, Klein-Flügge MC, Barry RL (2021) Structural and resting state functional connectivity beyond the cortex. NeuroImage 240:118379 Viessmann O, Polimeni JR (2021) High-resolution fMRI at 7Tesla: challenges, promises and recent developments for individual-focused fMRI studies. Curr Opin Behav Sci 40:96–104 Friston KJ, Harrison L, Penny W (2003) Dynamic causal modelling. NeuroImage 19:1273–1302 Friston KJ et al (2019) Dynamic causal modelling revisited. NeuroImage 199:730–744 Stephan KE, Penny WD, Moran RJ, den Ouden HEM, Daunizeau J, Friston KJ (2010) Ten simple rules for dynamic causal modeling. NeuroImage 49:3099–3109 Steward T et al (2022) A thalamo-centric neural signature for restructuring negative self-beliefs. Mol Psychiatry 27:1611–1617 Clark GI, Egan SJ (2015) The Socratic method in Cognitive Behavioural Therapy: A narrative review. Cognit Ther Res 39:863–879 Beck JS (2021) Cognitive behavior therapy: Basics and beyond, 3rd edn. The Guilford Press Cooper M, Cohen-Tovée E, Todd G, Wells A, Tovée M (1997) The Eating Disorder Belief Questionnaire: Preliminary development. Behav Res Ther 35:381–388 Ehring T, Zetsche U, Weidacker K, Wahl K, Schönfeld S, Ehlers A (2011) The Perseverative Thinking Questionnaire (PTQ): Validation of a content-independent measure of repetitive negative thinking. J Behav Ther Exp Psychiatry 42:225–232 Zeidman P et al (2019) A guide to group effective connectivity analysis, Part 1: First level analysis with DCM for fMRI. NeuroImage 200:174–190 Zeidman P et al (2019) A guide to group effective connectivity analysis, Part 2: Second level analysis with PEB. NeuroImage 200:12–25 Greaves MD, Novelli L, Razi A (2024) Structurally informed resting-state effective connectivity recapitulates cortical hierarchy. bioRxiv , 2024.2004.2003.587831 Sokolov AA, Zeidman P, Erb M, Ryvlin P, Pavlova MA, Friston KJ (2019) Linking structural and effective brain connectivity: structurally informed Parametric Empirical Bayes (si-PEB). Brain Struct Funct 224:205–217 Stephan KE, Tittgemeyer M, Knösche TR, Moran RJ, Friston KJ (2009) Tractography-based priors for dynamic causal models. NeuroImage 47:1628–1638 Friston KJ et al (2016) Bayesian model reduction and empirical Bayes for group (DCM) studies. NeuroImage 128:413–431 Matsumoto M, Hikosaka O (2009) Representation of negative motivational value in the primate lateral habenula. Nat Neurosci 12:77–84 Steward T et al (2020) Dynamic neural interactions supporting the cognitive reappraisal of emotion. Cereb Cortex 00:1–13 Leech R, Sharp DJ (2014) The role of the posterior cingulate cortex in cognition and disease. Brain 137:12–32 Leech R, Kamourieh S, Beckmann CF, Sharp DJ (2011) Fractionating the Default Mode Network: Distinct contributions of the ventral and dorsal posterior cingulate cortex to cognitive control. J Neurosci 31:3217–3224 Foster BL, Koslov SR, Aponik-Gremillion L, Monko ME, Hayden BY, Heilbronner SR (2023) A tripartite view of the posterior cingulate cortex. Nat Rev Neurosci 24:173–189 Andrews-Hanna JR, Smallwood J, Spreng RN (2014) The default network and self-generated thought: Component processes, dynamic control, and clinical relevance. Ann N Y Acad Sci 1316:29–52 Menon V (2023) 20 years of the default mode network: A review and synthesis. Neuron 111:2469–2487 Davey CG, Pujol J, Harrison BJ (2016) Mapping the self in the brain's default mode network. NeuroImage 132:390–397 Agathos J et al (2023) Differential engagement of the posterior cingulate cortex during cognitive restructuring of negative self- and social beliefs. Soc Cogn Affect Neurosci 18 Baker PM, Mizumori SJY (2017) Control of behavioral flexibility by the lateral habenula. Pharmacol Biochem Behav 162:62–68 Orsini CA, Moorman DE, Young JW, Setlow B, Floresco SB (2015) Neural mechanisms regulating different forms of risk-related decision-making: Insights from animal models. Neurosci Biobehav Rev 58:147–167 Baker PM, Raynor SA, Francis NT, Mizumori SJY (2017) Lateral habenula integration of proactive and retroactive information mediates behavioral flexibility. Neuroscience 345:89–98 Stalnaker TA, Cooch NK, Schoenbaum G (2015) What the orbitofrontal cortex does not do. Nat Neurosci 18:620–627 Klein-Flügge MC, Bongioanni A, Rushworth MFS (2022) Medial and orbital frontal cortex in decision-making and flexible behavior. Neuron 110:2743–2770 Niv Y (2019) Learning task-state representations. Nat Neurosci 22:1544–1553 Rudebeck PH, Rich EL (2018) Orbitofrontal cortex. Curr Biol 28:R1083–R1088 Schuck Nicolas W, Cai Ming B, Wilson Robert C, Niv Y (2016) Human orbitofrontal cortex represents a cognitive map of state space. Neuron 91:1402–1412 Bromberg-Martin ES, Hikosaka O (2011) Lateral habenula neurons signal errors in the prediction of reward information. Nat Neurosci 14:1209–1216 Mizumori SJY, Baker PM (2017) The lateral habenula and adaptive behaviors. Trends Neurosci 40:481–493 Mathis V, Barbelivien A, Majchrzak M, Mathis C, Cassel J-C, Lecourtier L (2016) The lateral habenula as a relay of cortical information to process working memory. Cereb Cortex 27:5485–5495 Khalighinejad N, Garrett N, Priestley L, Lockwood P, Rushworth MFS (2021) A habenula-insular circuit encodes the willingness to act. Nat Commun 12:6329 Namboodiri VMK, Rodriguez-Romaguera J, Stuber GD (2016) The habenula. Curr Biol 26:R873–R877 Sachella TE et al (2022) A novel role for the lateral habenula in fear learning. Neuropsychopharmacology 47:1210–1219 Mathis V, Cosquer B, Avallone M, Cassel J-C, Lecourtier L (2015) Excitatory transmission to the lateral habenula Is critical for encoding and retrieval of spatial memory. Neuropsychopharmacology 40:2843–2851 Mathis V, Lecourtier L (2017) Role of the lateral habenula in memory through online processing of information. Pharmacol Biochem Behav 162:69–78 Sheehan DV et al (1998) The Mini-International Neuropsychiatric Interview (MINI): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 59:22–33 American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders: DSM-5, 5th edn. American Psychiatric Association Ince S et al (2023) Subcortical contributions to salience network functioning during negative emotional processing. NeuroImage 270:119964 Osmo F et al (2018) The Negative Core Beliefs Inventory: Development and psychometric properties. J Cogn Psychother 32:67–84 Moeller S et al (2010) Multiband multislice GE-EPI at 7 tesla, with 16-fold acceleration using partial parallel imaging with application to high spatial and temporal whole-brain fMRI. Magn Reson Med 63:1144–1153 Sun H et al (2020) Extracting more for less: Multi-echo MP2RAGE for simultaneous T1-weighted imaging, T1 mapping, mapping, SWI, and QSM from a single acquisition. Magn Reson Med 83:1178–1191 Marques JP, Kober T, Krueger G, van der Zwaag W, Van de Moortele P-F, Gruetter R (2010) MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field. NeuroImage 49:1271–1281 Wilke M (2012) An alternative approach towards assessing and accounting for individual motion in fMRI timeseries. NeuroImage 59:2062–2072 Ashburner J (2007) A fast diffeomorphic image registration algorithm. NeuroImage 38:95–113 Kasper L et al (2017) The PhysIO toolbox for modeling physiological noise in fMRI data. J Neurosci Methods 276:56–72 Reynaud O, Jorge J, Gruetter R, Marques JP, van der Zwaag W (2017) Influence of physiological noise on accelerated 2D and 3D resting state functional MRI data at 7 T. Magn Reson Med 78:888–896 Glover GH, Li TQ, Ress D (2000) Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med 44:162–167 Birn RM, Smith MA, Jones TB, Bandettini PA (2008) The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration. NeuroImage 40:644–654 Chang C, Cunningham JP, Glover GH (2009) Influence of heart rate on the BOLD signal: the cardiac response function. NeuroImage 44:857–869 Behzadi Y, Restom K, Liau J, Liu TT (2007) A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. NeuroImage 37:90–101 Chakravarty MM et al (2013) Performing label-fusion-based segmentation using multiple automatically generated templates. Hum Brain Mapp 34:2635–2654 Germann J et al (2020) Fully Automated Habenula Segmentation Provides Robust and Reliable Volume Estimation Across Large Magnetic Resonance Imaging Datasets, Suggesting Intriguing Developmental Trajectories in Psychiatric Disease. Biol Psychiatry Cogn Neurosci Neuroimaging 5:923–929 Ahumada-Galleguillos P, Lemus CG, Díaz E, Osorio-Reich M, Härtel S, Concha ML (2017) Directional asymmetry in the volume of the human habenula. Brain Struct Funct 222:1087–1092 Ranft K, Dobrowolny H, Krell D, Bielau H, Bogerts B, Bernstein HG (2010) Evidence for structural abnormalities of the human habenular complex in affective disorders but not in schizophrenia. Psychol Med 40:557–567 Nieto-Castanon A, Whitfield-Gabrieli S (2022) CONN functional connectivity toolbox: RRID SCR_009550 release 22 Olszowy W, Aston J, Rua C, Williams GB (2019) Accurate autocorrelation modeling substantially improves fMRI reliability. Nat Commun 10:1220 Friston K (2009) Causal modelling and brain connectivity in functional magnetic resonance imaging. PLoS Biol 7:e1000033 Penny WD, Stephan KE, Mechelli A, Friston KJ (2004) Comparing dynamic causal models. NeuroImage 22:1157–1172 Kahan J, Foltynie T, Understanding DCM (2013) Ten simple rules for the clinician. NeuroImage 83:542–549 Friston K, Penny W (2011) Post hoc Bayesian model selection. NeuroImage 56:2089–2099 Rosa MJ, Friston K, Penny W (2012) Post-hoc selection of dynamic causal models. J Neurosci Methods 208:66–78 Xia M, Wang J, He Y (2013) BrainNet Viewer: A network visualization tool for human brain connectomics. PLoS ONE 8:e68910 Additional Declarations There is NO Competing Interest. Supplementary Files HabenulaEffectiveConnectivityncommssourcedataFINAL.xlsx Source data HabenulaEffectiveConnectivitySupplementaryInformationv9FINAL.docx Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 07 May, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5634827","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":398226950,"identity":"320d550d-11a1-4174-89b3-6b0204a1bbf7","order_by":0,"name":"Po-Han Kung","email":"","orcid":"https://orcid.org/0000-0003-4583-8993","institution":"University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Po-Han","middleName":"","lastName":"Kung","suffix":""},{"id":398226951,"identity":"7f0a37b8-985f-49e1-8f6a-79c44085460f","order_by":1,"name":"Matthew Greaves","email":"","orcid":"https://orcid.org/0000-0002-3438-2874","institution":"University of Melbourne; Monash University","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Greaves","suffix":""},{"id":398226952,"identity":"5321e550-8c4a-4c48-978b-ccf6bde472fb","order_by":2,"name":"Eva Guerrero-Hreins","email":"","orcid":"https://orcid.org/0000-0002-4586-6874","institution":"University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Eva","middleName":"","lastName":"Guerrero-Hreins","suffix":""},{"id":398226953,"identity":"26f336d6-3744-493b-82a4-08ad1c0b8ca1","order_by":3,"name":"Ben Harrison","email":"","orcid":"","institution":"The University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Ben","middleName":"","lastName":"Harrison","suffix":""},{"id":398226954,"identity":"9c7098af-eefa-486a-8405-a3ca9af2dc05","order_by":4,"name":"Christopher Davey","email":"","orcid":"https://orcid.org/0000-0003-1431-3852","institution":"The University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"","lastName":"Davey","suffix":""},{"id":398226955,"identity":"ad277c7b-78a2-4ac9-8e38-a79d0be4adbe","order_by":5,"name":"Kim Felmingham","email":"","orcid":"","institution":"University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Kim","middleName":"","lastName":"Felmingham","suffix":""},{"id":398226956,"identity":"b083b37c-daef-4292-a9c0-49771d3dfb62","order_by":6,"name":"Holly Carey","email":"","orcid":"","institution":"University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Holly","middleName":"","lastName":"Carey","suffix":""},{"id":398226957,"identity":"5081688f-17f3-4fea-bd8e-97264e5b3f2d","order_by":7,"name":"Priya Sumithran","email":"","orcid":"","institution":"Monash University; Alfred Health","correspondingAuthor":false,"prefix":"","firstName":"Priya","middleName":"","lastName":"Sumithran","suffix":""},{"id":398226958,"identity":"aa1f4426-40b0-42c3-9d00-4d873a3d763a","order_by":8,"name":"Robyn Brown","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Robyn","middleName":"","lastName":"Brown","suffix":""},{"id":398226959,"identity":"66e37693-4f96-4baa-b8ad-589e9d62fa99","order_by":9,"name":"Bradford Moffat","email":"","orcid":"","institution":"The University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Bradford","middleName":"","lastName":"Moffat","suffix":""},{"id":398226960,"identity":"cdb45161-ab69-4379-8f02-f2e72064df81","order_by":10,"name":"Rebecca Glarin","email":"","orcid":"","institution":"University of Melbourne","correspondingAuthor":false,"prefix":"","firstName":"Rebecca","middleName":"","lastName":"Glarin","suffix":""},{"id":398226949,"identity":"470f5abb-6288-4b12-b4a2-83b1734820e8","order_by":11,"name":"Trevor Steward","email":"data:image/png;base64,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","orcid":"","institution":"University of Melbourne","correspondingAuthor":true,"prefix":"","firstName":"Trevor","middleName":"","lastName":"Steward","suffix":""}],"badges":[],"createdAt":"2024-12-13 02:30:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5634827/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5634827/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-59611-7","type":"published","date":"2025-05-07T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":73174833,"identity":"a75e3b22-80a1-4e94-87e2-c6a8f47cb48e","added_by":"auto","created_at":"2025-01-07 12:00:33","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":54233,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCognitive restructuring paradigm. a \u003c/strong\u003eIn each of the task blocks, participants were first shown a common negative self-cognition statement on the screen (e.g., “I am incompetent in the things I do”, “My value depends on my body shape”) for 4 seconds. Each block featured a unique statement. \u003cstrong\u003eb \u003c/strong\u003eNext, participants were given 9s to decide and select either to restructure or repeat the negative statement that was on the screen. Participants indicated their choice via an MRI-compatible button box, which moved the black cursor to the elected choice. A counter was presented under each option indicating the remaining number of blocks they could select the respective strategy, ensuring equal numbers of statements being restructured or repeated throughout the task. \u003cstrong\u003ec \u003c/strong\u003eOnce the decision period lapsed, participants were shown the same statement accompanied by an instruction to engage in their chosen strategy for 12 seconds (restructure = challenge condition; repeat = repeat condition). \u003cstrong\u003ed \u003c/strong\u003eA jittered fixation cross was then presented for an average of 6 seconds (rest condition) before the next block commenced. Of note, the replication sample group underwent an abbreviated version of this paradigm containing 16 blocks that did not include negative self-cognition statements about food and body image.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/51376da29279ca45d8155f13.png"},{"id":73174834,"identity":"2ec8f98a-6ac0-41ca-aded-6c684b306366","added_by":"auto","created_at":"2025-01-07 12:00:33","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":408424,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTask-based neural activation and construction of the DCM model space.\u003c/strong\u003e \u003cstrong\u003ea-b\u003c/strong\u003e The heatmaps display the general linear model (GLM) results of the cognitive restructuring fMRI paradigm (\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eFDR\u003c/em\u003e\u003c/sub\u003e\u0026lt;0.05, \u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003eE\u003c/em\u003e\u003c/sub\u003e≥10), with the colour bars representing the t-statistics of the single-sample t-test (one-tailed). The warm colour map shows brain regions with increased activity during the restructuring of negative self-cognitions compared to repeating (Challenge \u0026gt; Repeat). The cool colour map highlights structures showing increased activity during the repetition of negative cognitions versus restructuring (Repeat \u0026gt; Challenge). These results confirmed the engagement of the habenula in negative self-cognition processing and were used to inform DCM model node selection. \u003cstrong\u003ec \u003c/strong\u003eConsecutive coronal views of the neural activation results of the Repeat \u0026gt; Challenge contrast are presented on the MNI152 template to highlight the habenula cluster showing increased activity during the repeating of negative self-cognitions compared to restructuring. \u003cstrong\u003ed \u003c/strong\u003eThe line graph plots the group-level blood-oxygen level dependent (BOLD) response (GLM estimated) of the habenula region-of-interest across the key conditions and when averaged across task epochs (rest, repeat, rest, challenge). The habenula showed sustained activity during the repeating of negative self-cognitions (repeat condition) and evoked response during cognitive restructuring (challenge condition). \u003cstrong\u003ee \u003c/strong\u003eBased on the GLM results and past literature on habenula connectivity, the bilateral habenula, right pOFC, PCC, and the hippocampus were selected as model nodes for DCM analysis. \u003cstrong\u003ef \u003c/strong\u003eA DCM model centred on the habenula was constructed and estimated for each individual. The model assumed 1) bidirectional endogenous connections between the habenula and the other network regions (grey arrows), in addition to their self-inhibitory connectivity (not shown here); 2) driving input of the experimental stimuli into all network nodes (yellow arrows); and 3) modulatory effects of the restructuring and repetition of negative self-cognitions on the bidirectional connections between the habenula and the PCC, hippocampus, as well as the pOFC (dashed blue arrows). 3-D brain rendering were constructed in BrainNet Viewer with the MNI152 template brain\u003csup\u003e109\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDCM \u003c/em\u003edynamic causal model, \u003cem\u003edlPFC \u003c/em\u003edorsolateral prefrontal cortex, \u003cem\u003edmPFC \u003c/em\u003edorsomedial prefrontal cortex, \u003cem\u003eHC \u003c/em\u003ehippocampus, \u003cem\u003eL \u003c/em\u003eleft, \u003cem\u003ePCC \u003c/em\u003eposterior cingulate cortex, \u003cem\u003epOFC \u003c/em\u003eposterior orbitofrontal cortex, \u003cem\u003epSMA \u003c/em\u003epre-supplementary motor area, \u003cem\u003eR \u003c/em\u003eright,\u003cem\u003e vlPFC \u003c/em\u003eventrolateral prefrontal cortex.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/ede725228045f29fa294a7e3.png"},{"id":73175806,"identity":"e45e11bf-4dc9-4ef4-a871-229c922ce7e1","added_by":"auto","created_at":"2025-01-07 12:08:33","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":105931,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHabenula effective connectivity in the discovery sample. a \u003c/strong\u003eIntrinsic connectivity and task-induced modulation of connections between the habenula and the network nodes that demonstrated strong evidence (posterior probability \u0026gt;.95) for a non-zero group effect are illustrated on the MNI152 template brain with BrainNet Viewer\u003csup\u003e109\u003c/sup\u003e. Intrinsic connectivity is represented with solid arrows while dashed arrows depict modulatory effects by each of the task conditions-of-interest (i.e., the restructuring or repeating of negative self-cognitions). Red arrows represent excitatory intrinsic effective connectivity or positive modulatory effects. Blue arrows show inhibitory intrinsic connectivity or negative task-induced modulation. \u003cstrong\u003eb \u003c/strong\u003eTask-induced changes in habenula effective connectivity are plotted for connectivity associated with significant modulatory effects. The first column shows the average effective connectivity throughout the task for each pathway (intrinsic connectivity; A-matrix). This represents the context-independent influence from the habenula to the PCC (blue) and the pOFC (green). The second and third column represent the net effective connectivity of each pathway under the repeat and challenge conditions, respectively. This is calculated by adding or subtracting the modulatory effects from the corresponding intrinsic connectivity parameter (A-matrix + B-matrix) as the experimental input is mean-centred in our model. All connectivity estimates are in units of Hz denoting rate of change in neural activity in the input regions (e.g., PCC, pOFC) due to neural response of the output region (i.e., habenula).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHb \u003c/em\u003ehabenula, \u003cem\u003eHC \u003c/em\u003ehippocampus, \u003cem\u003eHz \u003c/em\u003ehertz, \u003cem\u003eL \u003c/em\u003eleft, \u003cem\u003ePCC \u003c/em\u003eposterior cingulate cortex, \u003cem\u003epOFC \u003c/em\u003eposterior orbitofrontal cortex, \u003cem\u003eR \u003c/em\u003eright.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/b942f9cd839954e477e93d6d.png"},{"id":73174839,"identity":"f140c74d-f6f0-4a91-be0b-159245ba5295","added_by":"auto","created_at":"2025-01-07 12:00:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":132046,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHabenula effective connectivity in the replication sample. a \u003c/strong\u003eIntrinsic and task-induced modulation of habenula connectivity with strong evidence (posterior probability \u0026gt;.95) for a non-zero group effect are displayed on the MNI152 template brain using BrainNet Viewer\u003csup\u003e109\u003c/sup\u003e. Solid arrows represent intrinsic connectivity while dashed arrows illustrate modulatory effects by the task conditions (i.e., the restructuring or repeating of negative self-cognitions). Red arrows show excitatory intrinsic connectivity or positive modulatory effects. Blue arrows indicate inhibitory intrinsic connectivity or negative task-induced modulation. \u003cstrong\u003eb\u003c/strong\u003e Changes in habenula effective connectivity are plotted for connectivity showing significant task-induced modulation. The first column quantifies the average effective connectivity throughout the task for each pathway (intrinsic connectivity; A-matrix), representing the context-independent influence from the habenula to the PCC (blue) and the pOFC (green). The second and third column show the net effective connectivity of each pathway under the repeat and challenge conditions. As the experimental input is mean-centred in our model, this is calculated as the intrinsic connectivity parameter plus the modulatory effect of the corresponding task condition (A-matrix + B-matrix). All connectivity estimates are in units of Hz denoting rate of change in neural activity in the input regions (e.g., PCC, pOFC) due to neural response of the output region (i.e., habenula). \u003cstrong\u003ec \u003c/strong\u003eThe bar graph depicts modulatory connectivity of the discovery model (grey) superimposed with the posterior expectation of the replication model (orange). The bars represent the Bayesian model-averaged (BMA) connectivity strength estimates of the corresponding network connection of the models, and the whiskers show the 95% confidence interval (CI) of the discovery model parameter estimates derived from the posterior covariance matrix (spm_plot_ci.m). An asterisk is placed above the bars for connectivity replicated across the models, which are identified based on the replication model connectivity with posterior expectations that are within the 95% CI of the discovery model estimate and surpass the posterior probability threshold (\u0026gt;.95). The positive modulatory effect of the restructuring of negative cognitions is consistently observed in the discovery and replication models.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCHAL \u003c/em\u003echallenge condition, \u003cem\u003eHb \u003c/em\u003ehabenula, \u003cem\u003eHC \u003c/em\u003ehippocampus, \u003cem\u003eHz \u003c/em\u003ehertz, \u003cem\u003eL \u003c/em\u003eleft, \u003cem\u003ePCC \u003c/em\u003eposterior cingulate cortex, \u003cem\u003epOFC \u003c/em\u003eposterior orbitofrontal cortex, \u003cem\u003eR \u003c/em\u003eright, \u003cem\u003eREP \u003c/em\u003erepeat condition.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/bb97d0a1e638f350d3c19ae4.png"},{"id":73174842,"identity":"b8cea9b3-b5de-45a2-819d-14583efe8831","added_by":"auto","created_at":"2025-01-07 12:00:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":216235,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGeneration and evaluation of individual habenula ROIs. a\u003c/strong\u003e The MAGeTbrain algorithm produced high-resolution individualised habenula masks (red outline) using the 7-Tesla anatomical images (0.75 mm isotropic). Three example habenula masks are presented here and displayed on their corresponding T1-weighted whole-brain image in the native space. \u003cstrong\u003eb \u003c/strong\u003eThe individualised habenula masks were normalised to the MNI space using the DARTEL flow fields produced during anatomical image pre-processing. These high-resolution normalised masks were used as reference label images with which we evaluated the functional resolution habenula masks. \u003cstrong\u003ec \u003c/strong\u003eTo create the habenula ROIs for functional analysis, each individual’s anatomical-resolution habenula mask were resampled to the functional resolution (2mm isotropic) and transformed to the standard space. An iterative volume optimisation procedure was developed to minimise the impact of down-sampling on the habenula mask. In brief, adjustment was made to the re-binarising threshold in each iteration to ensure that the resultant habenula ROI had a volume that is within 10% of the reference label image (i.e., normalised habenula mask in anatomical resolution) after the removal of CSF voxels. \u003cstrong\u003ed \u003c/strong\u003eThe subject-specific habenula ROIs in functional resolution were further validated via a rsFC analysis. Results of this analysis replicated the habenula functional connectivity pattern reported in past studies.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eACC \u003c/em\u003eanterior cingulate cortex, \u003cem\u003eCSF \u003c/em\u003ecerebrospinal fluid,\u003cem\u003e DARTEL \u003c/em\u003eDiffeomorphic Anatomical Registration Through Exponentiated Lie Algebra, \u003cem\u003eFSL\u003c/em\u003eFMRIB Software Library, \u003cem\u003eL \u003c/em\u003eleft, \u003cem\u003eMAGeT \u003c/em\u003eMultiple Automatically Generated Templates brain segmentation,\u003cem\u003e PCC \u003c/em\u003eposterior cingulate cortex, \u003cem\u003epSMA \u003c/em\u003epre-supplementary motor area, \u003cem\u003eR\u003c/em\u003e right, \u003cem\u003ersFC\u003c/em\u003e resting-state functional connectivity, \u003cem\u003eSPM\u003c/em\u003e Statistical Parametric Mapping, \u003cem\u003eVOI \u003c/em\u003evolume-of-interest, \u003cem\u003eVTA \u003c/em\u003eventral tegmental area.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/a151900b98741e9d257e86f9.png"},{"id":82235550,"identity":"cabca660-328c-4c78-a3ee-d458c112331f","added_by":"auto","created_at":"2025-05-08 07:05:37","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2264887,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/9624f887-afd3-463a-a03b-c89e34b48875.pdf"},{"id":73174832,"identity":"230730be-1d8d-423d-a2c0-74859c06dd2a","added_by":"auto","created_at":"2025-01-07 12:00:33","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":12411,"visible":true,"origin":"","legend":"Source data","description":"","filename":"HabenulaEffectiveConnectivityncommssourcedataFINAL.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/56d1a79486c9af497434530d.xlsx"},{"id":73174854,"identity":"b97309c2-6e27-4aeb-9edd-f299030d0fea","added_by":"auto","created_at":"2025-01-07 12:00:34","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2862995,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"HabenulaEffectiveConnectivitySupplementaryInformationv9FINAL.docx","url":"https://assets-eu.researchsquare.com/files/rs-5634827/v1/7a16691fa89b5cdd42fb3a66.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Habenula neural circuitry drives negative self-cognitions","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSelf-cognitions are thoughts and beliefs about how an individual perceives themselves, their attributes, as well as their relationship with others and the world\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. These are often derived from personal experiences, helping to form a cohesive narrative of \u0026ldquo;Who am I?\u0026rdquo; and \u0026ldquo;What am I like?\u0026rdquo; that defines the distinctly human phenomenon of having a sense of \u0026lsquo;self\u0026rsquo;\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Importantly, self-cognitions influence how one evaluates and interprets past events, responds to present circumstances, and predicts future situations\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. As such, self-cognitions are deeply intertwined with an individual\u0026rsquo;s affective experience and play a principal role in psychological wellbeing\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. For instance, persistent engagement with negative self-cognitions in the form of repetitive negative thinking has been shown to contribute to mental ill health\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. In contrast, the ability to restructure and update negative self-cognitions with more adaptive narratives can alleviate negative affect and act as a protective factor for mental wellbeing\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Despite their significance to mental wellbeing, the brain mechanisms supporting the higher-order processing of self-cognitions remain largely unexplored. Understanding the neural mechanisms of negative self-cognitions would provide valuable insight into the biological basis of maladaptive thinking patterns, such as rumination, which contribute to depression and anxiety disorders\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe habenula \u0026ndash; a pair of small midbrain nuclei adjacent to the posterior mediodorsal thalamus \u0026ndash; may play a role in encoding negative self-cognitions owing to its distinctive function in the processing of other negative stimuli\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Converging animal and human studies have shown that habenula activity increases in response to the omission of expected reward\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e and to the delivery of punishment\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Conversely, the majority of habenula neurons show reduced firing during unexpected reward receipt and when facing reward-predictive cues\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. The habenula\u0026rsquo;s unique function in signalling negative valence and non-reward events has led to it being recognised as the \u0026lsquo;anti-reward\u0026rsquo; centre of the brain\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. In particular, rodent models have shown that neurochemical activation of the habenula induces depression-like symptoms characterised by reduced mobility and sucrose preference, which can be alleviated via pharmacological inhibition of the habenula\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. These findings have been complemented by human studies reporting increased habenula volume and activity in depression\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, as well as intensified habenula activity in response to negative feedback during cognitive tasks\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe habenula\u0026rsquo;s contribution to shaping affective and behavioural responses to negative stimuli is likely underpinned by its extensive connectivity bridging the forebrain to midbrain monoamine systems\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Specifically, the habenula receives efferent projections from the medial prefrontal cortex (mPFC), the basal ganglia (e.g., globus pallidus), and the lateral hypothalamus, which supply information related to an individual\u0026rsquo;s motivational state\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. The habenula, in turn, modulates downstream neurotransmission to shape cognition and behaviour via bidirectional connections with the ventral tegmental area (VTA), substantia nigra compacta, and the raphe nucleus\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. In rats, synaptic potentiation of habenula neurons projecting to the VTA has been found to modulate learned helplessness\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and elevated habenula activity has been shown to induce depressive behaviours by reducing serotonin transmission from the dorsal raphe\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Moreover, signals transmitted from the habenula to the mPFC via VTA dopaminergic neurons mediates conditioned place aversion in rats\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Relatedly, interactions between habenula neurons and the anterior cingulate cortex have been shown to guide choice shifting in response to unrewarding outcomes in primates during a reversal learning task\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. However, it is unclear if such habenula-mediated functions extend to higher-order cognition, such as negative self-related cognitions.\u003c/p\u003e \u003cp\u003eAs the processing of negative self-cognitions is a uniquely human process, the involvement of the habenula in encoding and restructuring self-cognitions cannot be tested via animal models. To date, human neuroimaging studies have largely focused on habenula response to primary reward or punishment, such as electric shocks\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, or under task-free/resting-state conditions\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. In addition to reward processing regions, resting-state imaging studies have found that habenula activity is correlated with the activity of the orbitofrontal cortex (OFC), hippocampus, and posterior cingulate cortex (PCC) \u0026ndash; key regions in large-scale networks implicated in outcome valuation, memory functioning and self-referential cognition\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. However, a mechanistic account of the habenula\u0026rsquo;s influence over these regions to support valence attribution and negative self-related cognitions remains undefined. Furthermore, assessing habenula response using standard 3-Tesla functional MRI (fMRI) is challenging as limitations in signal contrast and spatial resolution hinder the accurate delineation of the habenula from nearby structures\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Ultra-high field (7T) MRI can overcome these challenges by providing the superior image resolution and signal-to-noise ratio required to map habenula function\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this work, we present the first study characterising the habenula\u0026rsquo;s involvement in the processing of negative self-cognitions. We sought (1) to characterise habenula activity during the encoding and restructuring of negative self-cognitions and (2) to map the directional influence between the habenula and regions implicated in negative self-cognition processing via dynamic causal modelling (DCM). Under a Bayesian framework, DCM uses a neurobiologically informed generative model to infer the causal excitatory and inhibitory effects brain regions have on one another (i.e., effective connectivity), as well as to classify how these interactions are modulated by experimental tasks\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. Given the habenula\u0026rsquo;s role in signalling negative valence, we hypothesised that habenula activity would increase during the encoding of negative self-cognitions, and that this would be heightened when participants repeat negative cognitions as opposed to restructuring them. We further hypothesised that both the repeating and restructuring of negative self-cognitions would positively modulate connectivity within our habenula-centric network. In addition, we examined the extent to which habenula connectivity during negative self-cognition processing was associated with participants\u0026rsquo; endorsement of negative cognitions, as well as their tendency to engage in repetitive negative thinking. We carried out our analyses using a discovery sample including 48 healthy participants, and tested the replicability of our obtained findings in an independent replication sample comprising 65 healthy participants.\u003c/p\u003e \u003cp\u003eUsing 7T fMRI, we demonstrate that the repetition of negative self-cognitions elicits heightened activity in the habenula compared to restructuring, alongside regions implicated in self-directed thinking (e.g., PCC), outcome valuation (e.g., OFC), and memory (e.g., hippocampus). DCM analyses in the discovery sample reveal that the habenula exerted an excitatory influence on the PCC during both the restructuring and repeating of negative cognitions. In contrast, restructuring negative self-cognitions is characterised by the habenula having an excitatory modulatory effect on the OFC. Our replication sample corroborates this excitatory effect from the habenula to the OFC during the restructuring of negative self-cognitions, providing novel and consistent evidence for the habenula\u0026rsquo;s involvement in processing negatively valanced self-cognitions that extend beyond its previously limited role in encoding primary punishment and reward.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eCognitive restructuring paradigm\u003c/h2\u003e \u003cp\u003eAll participants completed a novel block-design cognitive restructuring paradigm\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e while undergoing fMRI scanning (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e; detailed in Methods). Prior to scanning, participants received training on how to restructure negative self-cognitions using Socratic questioning techniques, such as logical reasoning and perspective shifting\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. At the start of each block, participants were presented with a commonly reported negative self-cognition statement\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e and given the option to restructure or to repeat each statement. For half of the task blocks, participants restructured the negative self-cognition statements (challenge condition) using previously taught Socratic questioning techniques\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. For the other half of the task blocks, the participants silently repeated the statement to themselves\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e without engaging in any conscious attempts to refute the negative self-cognitions (repeat condition). Each task block concluded with a fixation cross (rest condition) before the next block began. Prior to and following scanning, participants rated the extent to which they agreed with the presented negative self-cognition statements. Participants also completed the Perseverative Thinking Questionnaire\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e to assess repetitive negative thinking tendencies (reported in Supplementary Table\u0026nbsp;1).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eHabenula activity \u0026 effective connectivity during negative self-cognition processing\u003c/h3\u003e\n\u003cp\u003eMass-univariate general linear model (GLM) activation analysis revealed that the habenula had increased activity during the repeating of negative self-cognitions compared to restructuring, in addition to the right PCC, right hippocampus, and the right pOFC (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e; Supplementary Fig.\u0026nbsp;1 \u0026amp; Table\u0026nbsp;2). As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, habenula response increased during both the repeating and restructuring of negative self-cognitions relative to rest. Complete GLM activation results are reported in Supplementary Table\u0026nbsp;2.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eBased on these initial activation results (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;48) and past neuroimaging evidence\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, a DCM network including the habenula, right PCC, right hippocampus, and right pOFC as regions-of-interest (model nodes) was inverted for each participant to infer the modulatory influence of negative self-cognition processing on habenula effective connectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee-f)\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Our hypothesised network structure assumed the presentation of negative self-cognition statements as driving input into all regions-of-interest. To probe the effect of habenula activity on the neuronal response of other network regions, we modelled the habenula\u0026rsquo;s bidirectional pathways to and from the other network nodes, in addition to their self-connections. For each of these pathways, we estimated their (1) intrinsic connectivity, which represent the context-independent interaction between the network regions, i.e., average effective connectivity across the entire paradigm; and (2) the modulatory effects of repeating or restructuring negative self-cognitions on interregional effective connectivity. Group-level effects were summarised using Parametric Empirical Bayes (PEB)\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e and thresholded at posterior probability\u0026thinsp;\u0026gt;\u0026thinsp;.95. The strength of effectivity connectivity is represented as a partial derivative, measured in hertz (Hz), that quantifies the rate at which neuronal activity in one region changes with respect to neuronal activity in another region (intrinsic connectivity) or due to an experimental input (modulatory effect).\u003c/p\u003e \u003cp\u003eDCM inversion (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;45) and PEB revealed that both the repeating and restructuring of negative self-cognitions positively modulated the connectivity from the habenula to the PCC, such that the habenula exerted an excitatory effect on the PCC during both task conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-b). The habenula-to-pOFC pathway was positively modulated by the restructuring of negative self-cognitions, suggesting that the habenula upregulated pOFC activity during cognitive restructuring but not the repeat condition. With regards to intrinsic effective connectivity, the PCC and the pOFC had an excitatory influence on the habenula when averaged across the entire cognitive restructuring task (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea); whereas the habenula had an inhibitory effect on the activity of the PCC. Complete Bayesian model-averaged parameter estimates, including posterior expectation, posterior covariance and posterior probability are reported in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBayesian model-averaged DCM parameters for endogenous and modulatory connections in the discovery sample\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConnection\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEp\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCp\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePP\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEndogenous connections\u003csup\u003ea\u003c/sup\u003e (A-matrix)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; PCC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; Hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; pOFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePCC \u0026rarr; PCC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0017\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePCC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.99*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHippocampus \u0026rarr; Hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHippocampus \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.93\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epOFC \u0026rarr; pOFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epOFC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eModulatory connections\u003c/b\u003e\u003csup\u003e\u003cb\u003eb\u003c/b\u003e\u003c/sup\u003e \u003cb\u003e(B-matrix)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eChallenge (CHAL)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; PCC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; Hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; pOFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0052\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePCC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHippocampus \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt; -0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epOFC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRepeat (REP)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; PCC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0099\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; Hippocampus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt; -0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHabenula \u0026rarr; pOFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt; -0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePCC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHippocampus \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epOFC \u0026rarr; Habenula\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003csup\u003ea\u003c/sup\u003eEndogenous parameters reflect the average effective coupling between regions across experimental conditions (context-independent).\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003csup\u003eb\u003c/sup\u003eModulatory parameters reflect the changes in effective coupling between regions induced by cognitive reappraisal (content-dependent).\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e*Posterior probability (PP) exceeding .95 provides sufficient evidence for a non-zero group effect\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003e\u003cem\u003eCp\u003c/em\u003e posterior covariance, \u003cem\u003eEp\u003c/em\u003e posterior expectation, \u003cem\u003epOFC\u003c/em\u003e posterior orbitofrontal cortex, \u003cem\u003ePP\u003c/em\u003e posterior probability, \u003cem\u003ePCC\u003c/em\u003e posterior cingulate cortex.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTwo additional PEB models including participant\u0026rsquo;s negative self-cognition endorsement and PTQ total scores as covariates did not show sufficient evidence to suggest that neither negative self-cognition endorsement nor perseverative thinking modulated habenula network dynamics (see Supplementary Table\u0026nbsp;3).\u003c/p\u003e\n\u003ch3\u003eIndependent sample replication \u0026 randomised 5-fold validation\u003c/h3\u003e\n\u003cdiv class=\"Heading\"\u003eIndependent sample replication \u0026amp; randomised 5-fold validation\u003c/div\u003e \u003cp\u003eWe evaluated the out-of-sample validity of our effective connectivity findings using an independent replication dataset (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;56). Brain activation patterns in the replication sample during both the repeating and restructuring of negative self-cognitions were largely consistent with the discovery sample (Supplementary Fig.\u0026nbsp;2, 3 \u0026amp; Table\u0026nbsp;4). For the replication sample, the specification of both the DCM and group-level PEB model replicated the procedures used in the discovery model. However, we leveraged the Bayesian model-averaged group-level posterior distribution from the discovery model, specified by its mean and covariance, as an empirical prior distribution over the effective connectivity parameters of the replication PEB model\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. This allowed us to inform the inversion of the replication PEB model and test the discovery model parameters in the independent dataset. The use of informed priors capitalises on the inherent advantage of the empirical Bayesian framework to test the validity of our findings \u0026ndash; that is, whether the same effective connectivity architecture is replicated in the independent dataset, given prior knowledge on the network dynamics derived from the discovery sample\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. Here, we report the DCM results from this informed replication model and summarise the parameter estimates in Supplementary Table\u0026nbsp;5. We also include the results of the non-informed model in Supplementary Table\u0026nbsp;6.\u003c/p\u003e \u003cp\u003eThe positive modulatory effects of the restructuring of negative self-cognitions on the habenula-to-pOFC pathway were identified again in the replication model, such that the habenula exerted an excitatory influence on the pOFC during cognitive restructuring (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The modulatory effects associated with the habenula-to-PCC pathway during restructuring and repeating were positive, though they did not surpass the posterior probability\u0026thinsp;\u0026gt;\u0026thinsp;.95 threshold.\u003c/p\u003e\u003cp\u003eTo further assess whether these findings were influenced by individual participant variance, we conducted a randomised stratified 5-fold validation using the combined dataset (including both the discovery and replication groups) and examined the consistency of connectivity results across the validation subsamples. As reported in Supplementary Table\u0026nbsp;7, the subsamples were comparable on sex (χ\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;7.60, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003ebonf.\u0026minus;corrected\u003c/em\u003e\u003c/sub\u003e = .535), age (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e4,96\u003c/sub\u003e = 0.69, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003ebonf.\u0026minus;corrected\u003c/em\u003e\u003c/sub\u003e = 1.000), perseverative thinking (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e4,96\u003c/sub\u003e = 1.39, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003ebonf.\u0026minus;corrected\u003c/em\u003e\u003c/sub\u003e = 1.000), and endorsement of negative self-cognitions (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e4,96\u003c/sub\u003e = 0.20, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003ebonf.\u0026minus;corrected\u003c/em\u003e\u003c/sub\u003e = 1.000). Consistent with the replication model described above, the 5-fold validation group-level DCM models were furnished with the empirical prior distribution derived from the posterior distribution of the discovery model. Through this procedure, we found very strong evidence (posterior probability\u0026thinsp;\u0026gt;\u0026thinsp;.95) supporting the positive modulatory connectivity of the habenula-to-pOFC during the restructuring of negative self-cognitions in 4 out of the 5 subsamples (Supplementary Fig.\u0026nbsp;4). Additionally, very strong evidence (posterior probability\u0026thinsp;\u0026gt;\u0026thinsp;.95) of positive modulation of the habenula-to-PCC connection during the restructuring and repeating conditions were replicated in 3 out of the 5 subsamples, upholding the reliability of our connectivity results despite moderate levels of inter-sample variability.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we combined DCM and 7T fMRI to infer the role of a habenula-centric circuitry during the processing of negative self-cognitions. In line with our hypothesis, we observed increased habenula activity during the repeating of negative self-cognitions compared to restructuring. Using data from two independently acquired samples, as well as a randomised 5-fold validation, we identified a reliable network structure revealing excitatory effective connectivity from the habenula to the pOFC during the restructuring of negative self-cognitions. In the discovery sample, we identified excitatory effective connectivity from the habenula to the PCC during both the repeating and restructuring of negative self-cognitions. These findings provide novel insights into the habenula\u0026rsquo;s functional influence on key nodes of the default mode network and cognitive control network to support self-related higher-order cognitions in humans, thereby broadening our understanding of habenula function to encompass domains not limited to external primary reward or punishment.\u003c/p\u003e \u003cp\u003eAlthough both the restructure and repeat task conditions modulated habenula activity, it is worth noting that habenula response was heightened during the repeating of negative self-cognitions. As negative self-cognitions were not consciously restructured and reduced in intensity during the repeat condition, increased habenula activity may reflect the sustained signalling of negative valence induced by the repetition of negative self-cognition statements. This aligns with the understanding that habenula activity encodes the negative motivational value of external punishment or unrewarding outcomes to influence behavioural response\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e, suggesting parallel neural processes between the encoding of the aversiveness of negative self-cognitions and negative reward signalling. This valence-based information may then be integrated into higher-order cognition subserved by other cortical systems\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eOur model demonstrated that both the restructuring and repeating of negative self-cognitions positively modulated habenula-to-PCC connectivity, such that the habenula consistently exerted an excitatory influence on the PCC during engagement with negative self-cognitions. As a core node of the default mode network, the PCC has been posited to play a coordinating role in the flexible attentional switch between internal and external environments\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. Through interactions with the frontoparietal executive control regions and the salience network (e.g., anterior cingulate cortex, anterior insula), the PCC receives information regarding the personal relevance of the task at-hand to inform attentional resource allocation\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e, with sustained PCC activity facilitating more self-oriented cognition\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. Crosstalk between the habenula and the PCC may similarly allow the incorporation of valence and motivational value to tilt attentional balance towards self-referential processes when faced with negative self-related cognitions. This hypothesis accords with the view that connectivity between the habenula and the default mode network potentially reflects an integrative self-monitoring process, with the value of negative stimuli being signalled by the habenula\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Moreover, the PCC is integral to the generation of a unitary representation of the self and performs a gating function by which self-conceptualisation enters conscious awareness\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. This is particularly relevant during autobiographical memory recall where the PCC-mediated self-conceptualisation represents egocentric information associated with prior experiences\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. Via the habenula-to-PCC pathway, negative valence encoded in the habenula may be attributed to mental representations of the self during associations triggered by the internal recital of negative statements or during the conscious recollection of personal experiences required to refute them.\u003c/p\u003e \u003cp\u003eThe habenula exerted a distinct excitatory effect on the pOFC during the restructuring of negative self-cognitions \u0026ndash; an effect that was reliably detected in both the discovery and replication samples. Restructuring negative self-cognitions not only requires the sustaining of complex self-concepts, subserved by the default mode network\u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e, but also the manipulation of self-representations, which has been shown to involve frontostriatal valuation and cognitive control circuits\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. The observed excitatory connectivity from the habenula to the pOFC concurs with these reports and suggests a previously undescribed role of the habenula in shaping adaptive responses to negative cognitions. Besides negative valence, the habenula is known to be sensitive to trial-to-trial feedback for reward outcomes\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e, and to contribute to the flexible modification of action strategies in reaction to aversive outcomes\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e. These findings point to the habenula\u0026rsquo;s contribution to tracking the effectiveness of behavioural responses in relation to outcome expectations and feedback throughout changing contexts. In this regard, excitatory connectivity from the habenula to the pOFC may act as a pathway through which the expectation and outcome of the restructuring effort is transmitted from the midbrain to the prefrontal cortex.\u003c/p\u003e \u003cp\u003eOur findings are in accordance with the growing consensus that the OFC constructs and maintains a cognitive map defining the current task space, in which multiple sources of information relevant to decision-making (e.g., action-outcome value, emotion, memory) are synthesised\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. The recruitment of the OFC is necessary in situations requiring mental simulation or future inferences, where values and predictions of possible outcomes associated with each choice options need to be computed with continuously updated information\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e. During the cognitive restructuring paradigm used in this study, participants were presented with a different statement on each trial and could not rely on previously formulated arguments to restructure the negative self-cognitions. Rather, participants need to dynamically adjust their cognitive strategies and to conceive new rebuttals in response to different self-cognition statements. Each restructuring strategy may represent an alternative task state that necessitates OFC-mediated representation and outcome-value computation\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. This real-time evaluation likely incorporates the moment-by-moment feedback about the expected and actual effectiveness of the restructuring strategy encoded by the habenula\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e, a process potentially subserved by the restructuring-induced positive modulation of the habenula-to-pOFC excitatory connectivity.\u003c/p\u003e \u003cp\u003eMizumori and Baker\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e recently synthesised findings from animal models to hypothesise that the habenula integrates action-outcome valuation from the mPFC and information about the organism\u0026rsquo;s internal state from the subcortex (e.g., lateral hypothalamus, entopeduncular nucleus) to signal the effectiveness of behaviours in relation to a contextual goal\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e. They proposed that the habenula encodes the decision to continue or alter the current course of action while simultaneously relaying this information to the hippocampus and the mPFC where subsequent actions may be updated and evaluated\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e. How this model applies to humans remains unknown. Nevertheless, our current findings support the role of habenula-frontal cortex interactions in the adaptive processing of negative cognitions and significantly expand the evidence base for the human habenula's role in complex, higher-order cognitive processes\u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eContrary to our hypotheses, we could not find sufficient evidence to support modulations in connectivity between the habenula and the hippocampus during the processing of negative self-cognitions. Insights into habenula-hippocampus interaction during mnemonic processing has primarily been derived from animal models of conditioned fear\u003csup\u003e\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e\u003c/sup\u003e, spatial and working memory\u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e, \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e\u003c/sup\u003e. How the habenula may support memory processes in humans, such as episodic memory recall for the restructuring of negative self-cognitions based on personal history, and working memory to sustain this process, remains an open question to be explored using different paradigms.\u003c/p\u003e \u003cp\u003eWe note some limitations of this work. First, the current samples included healthy participants who tended to report low levels of negative self-cognition endorsement and perseverative thinking. Thus, a floor effect may have impacted our ability to detect meaningful relationships between individual endorsement of negative self-cognitions and habenula connectivity. Relatedly, the lack of model evidence for an association between the perseverative thinking and negative self-cognition measures and habenula connectivity precluded inferences on the behavioural implication of habenula connectivity variations in healthy individuals. While the present study focused on establishing a normative role of the habenula in negative self-cognition processing, future studies would benefit from extending this to populations with elevated levels of maladaptive cognitions (e.g., people experiencing psychopathology)\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Lastly, there were minor methodological differences between our discovery and replication datasets that may have contributed to variance in our connectivity results. Despite these dissimilarities, we observed consistent modulation of the habenula connectivity during the restructuring of negative self-cognitions across the samples, which upholds the reliability of these findings.\u003c/p\u003e \u003cp\u003eHere, we present the first evidence in humans demonstrating habenula involvement in the higher-order processing of negative self-cognitions. We showed that habenula activity was modulated by the repeating and restructuring of negative self-cognitions. Using DCM in two independent samples, our model elucidated the directional interplay between the habenula, PCC, and the OFC, which prospectively underpins negative self-conceptualisation and the value-guided restructuring of negative self-cognitions. These findings advance our current understanding of the habenula\u0026rsquo;s role in negative stimuli processing beyond primary reward and punishment to include abstract internal experiences. A mechanistic account of habenula functioning lays the foundation for future work examining neural vulnerabilities contributing to maladaptive thinking patterns and whether the habenula represents a treatment target to alleviate entrenched negative self-cognitions that do not respond to conventional psychotherapy alone.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eFor our discovery sample, we recruited 57 healthy adults from the community via online advertisements. Inclusion criteria included: 1) being between the age of 18 and 40 years; 2) having no MRI contraindications (e.g., pregnancy, metallic implants or claustrophobia); 3) willingness to comply with the scanning centre\u0026rsquo;s healthy and safety policies (e.g., received full course of SARS-CoV-2 vaccination); 4) fluency in written and verbal English; 5) being capable of complying with study instructions. Participants were excluded if they 1) had a diagnosis of any mental disorder at the time of study participation; (2) have a past history of eating disorders, psychotic disorders, obsessive-compulsive disorder, or bipolar disorders based on The Mini International Neuropsychiatric Interview (MINI, English version 7.0.2)\u003csup\u003e\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e\u003c/sup\u003e for the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5)\u003csup\u003e\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e\u003c/sup\u003e; 3) have been diagnosed with autism spectrum disorder; 4) have major hearing or sight difficulties; or 5) have a medical or neurological condition for which they are on medication.\u003c/p\u003e \u003cp\u003eFor our independent replication sample, we obtained MRI data from 83 healthy adults, which have been described in previous reports\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e, \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u003c/sup\u003e. In brief, participants were eligible if they 1) aged between 18 and 40 years; 2) did not meet criteria for any mental disorders as screened using the MINI; 3) had no MRI contraindications; and 4) were proficient in English and had normal or corrected-to-normal vision.\u003c/p\u003e \u003cp\u003e All participants provided written informed consent and attended one testing session at the Melbourne Brain Centre Imaging Unit (The University of Melbourne, Parkville, Victoria, Australia). This study was approved by the University of Melbourne Human Research Ethics Committee (HREC 22347, 2056265).\u003c/p\u003e \u003cp\u003eNine and 18 participants were initially excluded from the discovery and replication samples, respectively, due to: technical errors during MRI acquisition (discovery: 1, replication: 3), participant not completing or incorrectly completing the fMRI paradigm (discovery: 7, replication: 5), and excessive head motion (discovery: 1, replication: 10). Thus, 48 participants from the discovery and 65 participants from the replication samples were included in the GLM activation analysis. An additional 3 participants from the discovery and 9 participants from the replication samples were excluded from the DCM analysis due to a failure to extract a valid time series from the regions-of-interest, resulting in 45 discovery group participants and 56 replication group participants being included in the final DCM analysis.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSelf-report measures\u003c/h3\u003e\n\u003cp\u003e \u003cb\u003eDemographic information.\u003c/b\u003e Age, sex (i.e., assigned sex at-birth), and ethno-cultural group based on participant self-report are summarised in Supplementary Table\u0026nbsp;1.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePerseverative thinking questionnaire (PTQ)\u003c/b\u003e \u003csup\u003e \u003cb\u003e \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e \u003c/b\u003e \u003c/sup\u003e. The PTQ is a 15-item questionnaire assessing individuals\u0026rsquo; general propensity to engage in repetitive negative thinking and its impact. Each item is rated on a 4-point Likert scale (0\u0026thinsp;=\u0026thinsp;never; 4\u0026thinsp;=\u0026thinsp;almost always), with higher total scores indicating stronger perseverative thinking tendencies. The PTQ was administered to all eligible participants prior to the MRI scanning session.\u003c/p\u003e \u003cp\u003e \u003cb\u003eChallenging negative beliefs task questionnaire (CNBTQ).\u003c/b\u003e The CNBTQ forms part of the cognitive restructuring fMRI task (see section below for detailed task description) and was administered prior to scanning and following scanning. The CNBTQ has participants rate the extent to which they endorse the negative self-cognition statements presented during the cognitive restructuring paradigm on a 7-point Likert scale (1\u0026thinsp;=\u0026thinsp;strongly disagree; 7\u0026thinsp;=\u0026thinsp;strongly agree). The difference in pre- and post-task ratings was used to index task-invoked shifts in negative self-cognition endorsement. Participants in the replication sample were only asked to provide post-task ratings on the statements that were restructured during the task, whereas participants in the discovery sample provided post-task ratings to all statements.\u003c/p\u003e\n\u003ch3\u003eBehavioural analyses\u003c/h3\u003e\n\u003cp\u003eDemographic and behavioural variables were analysed using SPSS v.27 (IBMCorp., Armonk, NY). Changes in negative self-cognition endorsement were compared within-group using single sample t-tests (2-tailed), and between-groups using independent sample t-tests (2-tailed). Likewise, two-tailed independent sample t-tests were used to assess differences in demographics and behavioural variables across the samples, whereas two-tailed Mann-Whitney U tests were used when there was evidence for heteroscedasticity. One-way analysis of variance (ANOVA) was used when three or more subgroups were compared. Two-tailed Pearson's chi-squared tests were adopted to evaluate categorical variables between the groups. Bonferroni correction was applied to account for multiple comparisons.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eCognitive restructuring paradigm\u003c/h2\u003e \u003cp\u003eAs described in Steward et al.\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, prior to scanning, participants received training to restructure negative cognitions using Socratic questioning techniques that mimic those practised in cognitive psychotherapy, such as recalling personal experiences that refute negatively biased self-conceptions\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Once research staff verified that the participant could correctly carry out Socratic questioning with the goal to restructure negative self-cognitions, participants completed the paradigm during MRI scanning. The cognitive restructuring task comprised one run of 24 blocks (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). In each block, the participants were first shown a statement on screen of common negative self-cognitions about the self, food, and body image reported in the cognitive behavioural therapy and psychopathology literature (e.g., \u0026ldquo;I am incompetent in the things I do\u0026rdquo;, \u0026ldquo;My value depends on my body shape\u0026rdquo;; Supplementary Table\u0026nbsp;8)\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e\u003c/sup\u003e. Next, participants decided via a button press whether they would repeat or restructure the presented statement using pre-trained strategies. Participants were instructed to restructure and repeat an equal number of statements. After 9 seconds, the participants were shown the same statement and instructed to engage in their chosen strategy for 12 seconds (restructure\u0026thinsp;=\u0026thinsp;challenge condition; repeat\u0026thinsp;=\u0026thinsp;repeat condition). A fixation cross was then presented for 6 seconds (rest condition) before the next block began with a new statement. The cognitive restructuring paradigm used in the discovery and replication samples were identical, except that the replication group completed an abbreviated version containing one run of 16 blocks that did not include negative self-cognition statements about food and body image.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003efMRI image acquisition\u003c/h2\u003e \u003cp\u003eImaging for the discovery and replication datasets was conducted on a 7-Tesla research scanner (Siemens Healthcare, Erlangen, Germany) equipped with an 8Tx/32Rx and 1Tx/32Rx head coil, respectively (Nova Medical Inc., Wilmington, MA, USA). The functional sequence was consistent across the two samples and consisted of a multi-band (factor\u0026thinsp;=\u0026thinsp;6) and GRAPPA (R\u0026thinsp;=\u0026thinsp;2) accelerated GE-EPI sequence\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e\u003c/sup\u003e in the steady state (TR\u0026thinsp;=\u0026thinsp;800 ms; TE\u0026thinsp;=\u0026thinsp;22.2 ms; pulse angle\u0026thinsp;=\u0026thinsp;45\u0026deg;; field of view\u0026thinsp;=\u0026thinsp;20.8 cm; acquisition matrix\u0026thinsp;=\u0026thinsp;130 \u0026times; 130-pixel; slice thickness\u0026thinsp;=\u0026thinsp;1.6 mm, no gap). Eighty-four interleaved axial slices were acquired along the anterior-posterior commissure line. In total, 946 and 628 whole-brain EPI volumes were acquired in a single run for the discovery and replication datasets, respectively, corresponding to approximately 12.6 and 8.4 minutes for the two task versions. High-resolution T1-weighted anatomical images were acquired for functional time-series co-registration and individualised habenula segmentation (detailed in section \u003cem\u003eHabenula segmentation \u0026amp; region-of-interest validation\u003c/em\u003e) using the first echo of a multi-echo Magnetization Prepared 2 Rapid Acquisition Gradient Echoes sequence (ME-MP2RAGE)\u003csup\u003e\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e for the discovery dataset (224 interleaved axial slices; TR\u0026thinsp;=\u0026thinsp;4500 ms; TE\u0026thinsp;=\u0026thinsp;2.21/4.21/6.15/8.14 ms; inversion time\u0026thinsp;=\u0026thinsp;700/2700 ms; flip angle\u0026thinsp;=\u0026thinsp;6/7\u0026deg;; field of view\u0026thinsp;=\u0026thinsp;24 cm; acquisition matrix\u0026thinsp;=\u0026thinsp;320 \u0026times; 320-pixel; slice thickness\u0026thinsp;=\u0026thinsp;0.75 mm, no gap), and a single-echo MP2RAGE\u003csup\u003e\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e\u003c/sup\u003e sequence for the replication dataset (224 interleaved sagittal slices; TR\u0026thinsp;=\u0026thinsp;5000 ms; TE\u0026thinsp;=\u0026thinsp;2.04 ms; inversion time\u0026thinsp;=\u0026thinsp;700/2700 ms; flip angle\u0026thinsp;=\u0026thinsp;4/5\u0026deg;; field of view\u0026thinsp;=\u0026thinsp;24 cm; acquisition matrix\u0026thinsp;=\u0026thinsp;320 \u0026times; 320-pixel; slice thickness\u0026thinsp;=\u0026thinsp;0.75 mm, no gap). Standard foam pads were used for all participants to minimize head movement during scanning. Respiratory and cardiac recordings were sampled at 50 Hz using a respiratory belt and pulse-oximeter. Information derived from these recordings were used for physiological noise correction during image pre-processing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003efMRI image pre-processing\u003c/h2\u003e \u003cp\u003eImaging data was pre-processed with Statistical Parametric Mapping 12 (SPM12, v7771; Wellcome Trust Centre for Neuroimaging, London) within the MATLAB 2023a environment (The MathWorks Inc., Natick, MA). Each participant\u0026rsquo;s functional time series was realigned to the mean image to correct for movement during the scan, and all images were resampled using 4th Degree B-Spline interpolation. Individual head motion was assessed with motion fingerprint\u003csup\u003e\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e\u003c/sup\u003e. Participants were excluded if they had a mean total displacement over 1.6 mm (one isotropic voxel size) and/or a maximum scan-to-scan displacement exceeding 2 mm. We imposed a stringent censoring criterion to minimise the effect of motion on signal distortion from the habenula. Each participant\u0026rsquo;s anatomical T1 image was co-registered to their mean functional image, segmented and normalized to the International Consortium for Brain Mapping (ICBM) European brain template using the unified segmentation approach plus Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL)\u003csup\u003e\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e\u003c/sup\u003e. Lastly, the functional images were spatially normalised with the DARTEL flow fields and smoothed with a 2 mm full width at half maximum (FWHM) Gaussian kernel to preserve spatial specificity.\u003c/p\u003e \u003cp\u003eCardiac and respiratory recordings were modelled using the PhysIO toolbox\u003csup\u003e\u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e\u003c/sup\u003e to account for physiological noise\u003csup\u003e\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e\u003c/sup\u003e. Specifically, the Retrospective Image-based Correction function (RETROICOR)\u003csup\u003e\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e\u003c/sup\u003e, respiratory response function (RRF)\u003csup\u003e\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e\u003c/sup\u003e and cardiac response function (CRF)\u003csup\u003e\u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e\u003c/sup\u003e were incorporated to correct for both periodic and variable effects of heartbeat and breathing, as well as their interaction, on BOLD signal. Anatomical component correction (aCompCor)\u003csup\u003e\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e\u003c/sup\u003e was applied to account for non-neural signal, whereby the mean and first principal components of the time series originating from the white matter (WM) and cerebrospinal fluid (CSF) were extracted using individualised DARTEL tissue maps.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eHabenula segmentation \u0026amp; region-of-interest (ROI) validation\u003c/h2\u003e \u003cp\u003eIndividual-specific habenula mask was generated in the native space using each participant\u0026rsquo;s high-resolution anatomical image (0.75 mm isotropic) via a validated, fully automated segmentation algorithm (Multiple Automatically Generated Templates Brain Segmentation Algorithm; MAGeTbrain)\u003csup\u003e\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e\u003c/sup\u003e. A schematic of our image preparation pipeline is included in the Supplementary Fig.\u0026nbsp;5. Using participant anatomical images and a set of habenula atlases as input, MAGeTbrain first propagates the segmentation of the atlas images to each subject image via a multi-stage image registration procedure to yield a large number of candidate segmentations for each individual (5 atlases \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\times\\:\\:\\)\u003c/span\u003e\u003c/span\u003e21 templates\u0026thinsp;=\u0026thinsp;105). Next, the candidate segmentations are fused via majority vote at each voxel, whereby the most frequently occurring label (habenula vs. non-habenula) was adopted in the final output segmentation. This approach reduces potential biases related to atlas or rater inconsistency, allows for neuroanatomical variability\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e, and has been shown to produce reliable habenula volume estimates across diverse populations and image acquisition parameters\u003csup\u003e\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe MAGeTbrain algorithm produced a binary anatomical-resolution habenula mask in the native space, which we then used to generate functional-resolution habenula ROIs in the standard space for fMRI analyses. We employed an iterative volume optimisation strategy adapted from Ely et al.\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e to minimise the impact of interpolation during image down-sampling, as well as the increased susceptibility to non-neural noise at a reduced resolution, to produce the most feasibly precise representation of each individual\u0026rsquo;s habenula ROI in the functional space (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e):\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003e1) Each anatomical habenula mask produced by MAGeTbrain, as well as their corresponding DARTEL segmented CSF mask, were resampled with trilinear interpolation to the functional resolution (2 mm isotropic) based on the mean functional image (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea), normalised to the standard space using their corresponding DARTEL flowfields, and binarised at a conventional threshold of 0.2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec), which resulted in habenula and CSF masks showing good spatial consistency with the original anatomical masks.\u003c/p\u003e \u003cp\u003e2) To maximally reduce signal contamination from the CSF, voxels in the resampled habenula mask that overlap with the resampled CSF mask (binarising threshold\u0026thinsp;=\u0026thinsp;0.2) were removed.\u003c/p\u003e \u003cp\u003e3) The volume of the adjusted habenula mask from Step 2 was compared to the reference habenula segmentation image (i.e., subject-specific normalised habenula mask in anatomical resolution; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003e4a) If the volume of the adjusted habenula mask from Step 2 fell within \u0026plusmn;\u0026thinsp;10% of the reference image, it was accepted for use as an ROI in subsequent analyses. The volume criterion was designed to optimally approximate our functional habenula ROI to the high-resolution segmentation, while taking into consideration individual anatomical variability.\u003c/p\u003e \u003cp\u003e4b) If the volume was above \u0026plusmn;\u0026thinsp;10% that of the reference image, it was rejected and Step 1 was repeated with the binarising threshold adjusted upwards or downwards by 0.01, producing a slightly smaller or larger habenula mask. Steps 2\u0026ndash;3 were then repeated. This process was iterated until the habenula mask at functional resolution was accepted (Step 4a).\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eThis iterative volume optimisation generated individualised habenula masks with excellent spatial specificity confirmed via visual inspection and volume estimates comparable with previous reports (i.e., 30\u0026ndash;60 mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e combined; discovery sample: Mean\u0026thinsp;=\u0026thinsp;60.93 mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, 95% CI = [57.72, 64.13]; replication sample: Mean\u0026thinsp;=\u0026thinsp;48.40 mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, 95% CI = [46.35, 50.44])\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e, \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe further validated the functional specificity of our subject-specific habenula ROI via a resting-state seed-to-whole brain functional connectivity analysis conducted in CONN22a\u003csup\u003e\u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e\u003c/sup\u003e. Please see Supplementary Methods for detailed information on image acquisition, pre-processing, and analysis. Using the individualised habenula ROI as a seed, we replicated the connectivity patterns reported in past studies at standard and ultra-high field strengths (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed)\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, highlighting habenula connectivity with structures such as the ventral tegmental area, thalamus, insula, anterior through to the posterior cingulate cortex, and the supplementary motor area.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eGeneral Linear Modelling (GLM) analysis\u003c/h2\u003e \u003cp\u003eSingle-subject (first-level) contrast images were estimated for Challenge\u0026thinsp;\u0026gt;\u0026thinsp;Repeat and Repeat\u0026thinsp;\u0026gt;\u0026thinsp;Challenge (i.e., the entire 12s when participants either restructured or repeated the negative self-cognition statements) to characterise changes in brain activation associated with the restructuring and repeating of negative self-cognitions, respectively. Each participant\u0026rsquo;s pre-processed timeseries, 6 realignment parameters for motion artifacts, and 24 physiological noise regressors (6 cardiac, 8 respiratory, 4 cardiac \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\times\\:\\:\\)\u003c/span\u003e\u003c/span\u003erespiratory regressors, 1 RRF, 1 CRF, top principal components and mean time series of the CSF and WM) were included in the GLM analysis, with the onset times for each condition event specified and convolved with the SPM canonical hemodynamic response function (HRF). Low-frequency fluctuation was high-pass filtered at 128Hz. The FAST method was used to estimate temporal autocorrelation resulting from our sub-second TR\u003csup\u003e\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e\u003c/sup\u003e. The first-level contrast images were brought forward to a group-level GLM (one-sample t-test, one-tailed). All GLM analyses were thresholded at a whole brain, false discovery rate (FDR) corrected \u003cem\u003eP\u003c/em\u003e\u003csub\u003eFDR\u003c/sub\u003e \u0026lt; 0.05, \u003cem\u003eK\u003c/em\u003e\u003csub\u003eE\u003c/sub\u003e \u0026ge; 10 voxels.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eDynamic Causal Modelling (DCM)\u003c/h2\u003e \u003cp\u003eDCM is a Bayesian framework used to infer the directional and causal influence that brain regions exert on one another (i.e., effective connectivity)\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. DCM enables inferences to be made by simulating neuroimaging timeseries via a generative model that is grounded in empirical knowledge of neuronal processes and how they translate to observable responses (e.g., BOLD signal)\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e\u003c/sup\u003e. Within a set of researcher-specified hypotheses regarding network structure, DCM estimates connectivity parameters through model inversion in a manner that optimises the trade-off between model fit to the observed data and model complexity \u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. This balance is reflected by free energy, which approximates the (log) model evidence used for model comparison and hypothesis testing between competing models\u003csup\u003e\u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e\u003c/sup\u003e. The model with the most positive free energy reflects a parsimonious and physiologically plausible account of the dynamic interactions between unobserved neuronal populations\u003csup\u003e\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e\u003cb\u003eModel space \u0026amp; timeseries extraction.\u003c/b\u003e Our DCM network included the bilateral habenula as a single ROI, as well as brain regions that were co-activated during Repeat\u0026thinsp;\u0026gt;\u0026thinsp;Challenge that have structural/functional connectivity with the habenula based on prior literature\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Specifically, the right PCC, hippocampus, and pOFC were included due to their central role in self-referential thinking, episodic memory and learning, as well as adaptive processes that are engaged during negative self-cognition processing. Representative time series (volume-of-interest) were extracted from these areas for each of the subjects following published guidelines\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe habenula ROI was delineated with the individualised mask described above. Whereas all other network region ROIs were defined as a 4 mm radius sphere centred around the individual neural activation maxima under the Repeat\u0026thinsp;\u0026gt;\u0026thinsp;Challenge contrast, constrained to be within 8 mm from each dataset\u0026rsquo;s group peak (Supplementary Table\u0026nbsp;9). Volume-of-interest from each of these regions were calculated using SPM as the principal eigenvariate of all voxels within the respective ROI that showed meaningful activation for the task contrast (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05, uncorrected) at the single subject level. In the case when an ROI contains no voxel surpassing the preset threshold, the statistical threshold incrementally relaxed to \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.5 (uncorrected) until a peak coordinate can be identified. This approach ensured minimal exclusion of participant data from our analyses as subjects lacking strong responses in a brain region or experimental condition may nevertheless provide useful information about other regions, conditions, and individual variability\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Of note, the contrast image used for VOI extraction was denoised with the motion and physiological nuisance regressors described in the General Linear Model section above, except the aCompCor components. This adjustment took into consideration the explicit removal of the CSF voxels from our individualised habenula mask and habenula\u0026rsquo;s high WM density to prevent overcorrection of the extracted timeseries, while minimising non-neural noise. Individual timeseries were additionally pre-whitened to mitigate serial correlations, high-pass filtered, and nuisance effects not covered by the Effects of Interest F-contrast are regressed out of the timeseries (i.e. \u0026lsquo;adjusted\u0026rsquo; to the F-contrast). Using this procedure, we extracted a complete set of VOIs for 45 and 56 individuals from the discovery and replication cohorts, respectively, which were then included in the connectivity analyses.\u003c/p\u003e \u003cp\u003e \u003cb\u003eModel specification \u0026amp; estimation.\u003c/b\u003e Our model was specified with SPM DCM 12.5 and featured 1) the endogenous connections between and within each target region (A-matrix), 2) the modulatory effect of the task conditions on inter-region connectivity (B-matrix), as well as 3) the driving influence of the task stimuli (C-matrix). All experimental input was mean centred to aid parameter interpretability. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, the full model assumed bidirectional endogenous connections of the habenula to and from the other regions in addition to their self-inhibitory connectivity, modelling the average connectivity parameters throughout the experiment that is independent of condition effects. We allowed the challenge and repeat conditions to modulate the connectivity between the habenula and other network nodes to characterise changes to habenula connectivity during negative self-cognition processing. Lastly, the aggregate of the 4s of statement presentation and 12s of restructuring/repeating negative self-cognitions was specified as the driving input to all nodes, reflecting the engagement of these regions during the exposure to negative self-cognition statements. This full model was estimated and evaluated for each subject, yielding individual posterior connectivity parameter estimates and their posterior probability.\u003c/p\u003e \u003cp\u003e \u003cb\u003eParametric empirical Bayes.\u003c/b\u003e Next, a group-level summary of the connectivity parameters was obtained via Parametric empirical Bayes (PEB)\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. PEB is a hierarchical model that incorporates both the subject-level parameter estimates and their uncertainty (i.e., posterior covariance) to the group level. Contrary to the standard summary statistics approach, the PEB framework effectively downweights data with noise and uncertain individual estimates to produce more reliable population connectivity estimates\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. Additionally, each level of the PEB hierarchy serves as a prior on the estimates of the level below it. This can improve the precision of individual parameter estimates by incorporating knowledge around task effects garnered from the cohort.\u003c/p\u003e \u003cp\u003eOur PEB model was designed to investigate the between-subject commonalities in connectivity parameters for each of the samples. The design matrix included an intercept term (single column of ones) denoting the overall mean connectivity\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Parameters from both A- and B-matrices were summarised in this model to account for potential conditional dependency. Generic prior distribution were adopted for the discovery model with no assumptions around the strength and variance of network connectivity\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Once this full PEB model was inverted, Bayesian Model Reduction (BMR) was used to search and compare the relative evidence of possible reduced models, iteratively pruning parameters that do not contribute to an increase in model evidence\u003csup\u003e\u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e107\u003c/span\u003e, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e108\u003c/span\u003e\u003c/sup\u003e. Bayesian Model Averaging (BMA) was then used to aggregate the parameters of the reduced models, weighted by the corresponding model\u0026rsquo;s posterior probability, to provide the final group summary\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. The BMA parameters were thresholded at posterior probability\u0026thinsp;\u0026gt;\u0026thinsp;.95, indicating sufficient evidence for a non-zero group effect.\u003c/p\u003e \u003cp\u003eThe two PEB models investigating the effect of negative self-cognition endorsement and repetitive negative thinking tendencies on habenula connectivity included the individual pre-task CNBTQ and PTQ total scores, respectively, as covariate regressors in addition to the intercept term. As all regressors were mean-centred, the between-subject effects could be quantified as the addition to or subtraction from the overall mean connectivity estimates\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFor the replication and 5-fold validation models, the generic prior distribution was substituted in the PEB model with the posterior distribution derived from the discovery model BMA summary. The discovery model posteriors served as 3rd -level empirical priors to constrain the 2nd -level (group-level) estimates of the replication and 5-fold validation models. As the result of BMR applied during the discovery model estimation, only connectivity with a non-zero posterior probability were evaluated in the replication and validation PEB models. This approach allows the testing of the discovery model parameters on the independent dataset and effectively incorporates empirically derived beliefs around network dynamics and their degree of uncertainty to refine the estimation of connectivity parameters in new populations\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e "},{"header":"Declarations","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eDeidentified effective connectivity data for this study are publicly available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/pohankung/NegativeBeliefs_Habenula_DCM\u003c/span\u003e\u003cspan address=\"https://github.com/pohankung/NegativeBeliefs_Habenula_DCM\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Source data for Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb are provided with the submission.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eScripts used to generate the main results and figures of this study are available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/pohankung/NegativeBeliefs_Habenula_DCM\u003c/span\u003e\u003cspan address=\"https://github.com/pohankung/NegativeBeliefs_Habenula_DCM\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Custom code was written in MATLAB 2023a. Statistical Parametric Mapping 12 (SPM12) and FMRIB Software Library (FSL) 6.0.6.5 were used for MRI processing.\u003c/p\u003e \u003c/div\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eP.-H.K., B.J.H., and T.S. conceived the general concept of this study, designed the experiment, and developed the model with input from M.D.G. and E.G.-H. P.-H.K., E.G.-H., B.J.H., K.L.F., H.C., P.S., R.M.B., B.A.M., R.K.G., and T.S. aided with data collection and the crafting of the imaging protocol. P.-H.K. and T.S. conducted data analysis with support from M.D.G. B.J.H., C.G.D., K.L.F., P.S., R.M.B., and T.S. provided supervision throughout the study. P.-H.K. and T.S. wrote the original draft of the manuscript. All authors reviewed and approved the final edit of this manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eWe thank James Agathos, Carly Beveridge, Lieselotte Claes, Yingliang Dai, Elizabeth Haris, Sevil Ince, Amy Nielson, Mia O\u0026rsquo;Shea, Tudor Sava, Braden Thai, and Andong Zhou for their contribution to data collection. We acknowledge the technical and scientific assistance of the Australian National Imaging Facility \u0026ndash; a National Collaborative Research Infrastructure Strategy (NCRIS) capability at the Melbourne Brain Centre Imaging Unit (MBCIU), The University of Melbourne. The multiband fMRI sequence was generously supported by a research collaboration agreement with CMRR, The University of Minnesota. Siemens Healthineers (Germany) provided the MP2RAGE sequence. This study was supported by the National Health and Medical Research Council of Australia (NHMRC)/Medical Research Future Fund (MRFF) Investigator Grant (MRF1193736), a Brain \u0026amp; Behaviour Research Foundation (BBRF) Young Investigator Grant and a University of Melbourne McKenzie Fellowship to T.S. Collection of the replication dataset was supported by a NHMRC Project Grant (1161897) to B.J.H. and an NHMRC Program Grant (1073041) to K.L.F.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLeary MR, Tangney JP (2012) The self as an organizing construct in the behavioral and social sciences. In: \u003cem\u003eHandbook of self and identity\u003c/em\u003e (ed^(eds Leary MR, Tangney JP). 2nd edn. The Guilford Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGallagher S (2000) Philosophical conceptions of the self: implications for cognitive science. Trends Cogn Sci 4:14\u0026ndash;21\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSegal ZV (1988) Appraisal of the self-schema construct in cognitive models of depression. Psychol Bull 103:147\u0026ndash;162\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoven M, Luigjes J, Denys D, Rouault M, van Holst RJ (2023) How do confidence and self-beliefs relate in psychopathology: A transdiagnostic approach. Nat Mental Health 1:337\u0026ndash;345\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEhring T, Watkins ER (2008) Repetitive negative thinking as a transdiagnostic process. Int J Cogn Ther 1:192\u0026ndash;205\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWahl K et al (2019) Is repetitive negative thinking a transdiagnostic process? A comparison of key processes of RNT in depression, generalized anxiety disorder, obsessive-compulsive disorder, and community controls. J Behav Ther Exp Psychiatry 64:45\u0026ndash;53\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBraun JD, Strunk DR, Sasso KE, Cooper AA (2015) Therapist use of Socratic questioning predicts session-to-session symptom change in cognitive therapy for depression. Behav Res Ther 70:32\u0026ndash;37\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEzawa ID, Hollon SD (2023) Cognitive restructuring and psychotherapy outcome: A meta-analytic review. Psychotherapy 60:396\u0026ndash;406\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcEvoy PM, Watson H, Watkins ER, Nathan P (2013) The relationship between worry, rumination, and comorbidity: Evidence for repetitive negative thinking as a transdiagnostic construct. J Affect Disord 151:313\u0026ndash;320\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatkins ER, Roberts H (2020) Reflecting on rumination: Consequences, causes, mechanisms and treatment of rumination. Behav Res Ther 127:103573\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMondoloni S, Mameli M, Congiu M (2022) Reward and aversion encoding in the lateral habenula for innate and learned behaviours. Transl Psychiatry 12:3\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHikosaka O (2010) The habenula: From stress evasion to value-based decision-making. Nat Rev Neurosci 11:503\u0026ndash;513\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShabel SJ, Wang C, Monk B, Aronson S, Malinow R (2019) Stress transforms lateral habenula reward responses into punishment signals. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e 116, 12488\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSosa R, Mata-Lu\u0026eacute;vanos J, Buenrostro-J\u0026aacute;uregui M (2021) The role of the lateral habenula in inhibitory learning from reward omission. \u003cem\u003eeneuro\u003c/em\u003e 8, ENEURO.0016-0021.2021\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson RP et al (2014) The habenula encodes negative motivational value associated with primary punishment in humans. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e 111, 11858\u0026ndash;11863\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatsumoto M, Hikosaka O (2007) Lateral habenula as a source of negative reward signals in dopamine neurons. Nature 447:1111\u0026ndash;1115\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeidacker K, Kim S-G, Nord CL, Rua C, Rodgers CT, Voon V (2021) Avoiding monetary loss: A human habenula functional MRI ultra-high field study. Cortex 142:62\u0026ndash;73\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbles JL, Park K, Iba\u0026ntilde;ez\u0026ndash;Tallon I (2023) Understanding the habenula: A major node in circuits regulating emotion and motivation. Pharmacol Res 190:106734\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoulos L-J, Darcq E, Kieffer BL (2017) Translating the Habenula - From Rodents to Humans. Biol Psychiatry 81:296\u0026ndash;305\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmidt FM et al (2017) Habenula volume increases with disease severity in unmedicated major depressive disorder as revealed by 7T MRI. Eur Arch Psychiatry Clin Neurosci 267:107\u0026ndash;115\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eProulx CD, Hikosaka O, Malinow R (2014) Reward processing by the lateral habenula in normal and depressive behaviors. Nat Neurosci 17:1146\u0026ndash;1152\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLecourtier L, Kelly PH (2007) A conductor hidden in the orchestra? Role of the habenular complex in monoamine transmission and cognition. Neurosci Biobehav Rev 31:658\u0026ndash;672\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFakhoury M (2017) The habenula in psychiatric disorders: More than three decades of translational investigation. Neurosci Biobehav Rev 83:721\u0026ndash;735\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu H, Cui Y, Yang Y (2020) Circuits and functions of the lateral habenula in health and in disease. Nat Rev Neurosci 21:277\u0026ndash;295\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStamatakis AM, Van Swieten M, Basiri ML, Blair GA, Kantak P, Stuber GD (2016) Lateral hypothalamic area glutamatergic neurons and their projections to the lateral habenula regulate feeding and reward. J Neurosci 36:302\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson RP, Drevets WC, Roiser JP (2013) Defining the habenula in human neuroimaging studies. NeuroImage 64:722\u0026ndash;727\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHennigan K, D'Ardenne K, McClure SM (2015) Distinct midbrain and habenula pathways are involved in processing aversive events in humans. J Neurosci 35:198\u0026ndash;208\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStamatakis Alice M et al (2013) A unique population of ventral tegmental area neurons inhibits the lateral habenula to promote reward. Neuron 80:1039\u0026ndash;1053\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi B et al (2011) Synaptic potentiation onto habenula neurons in the learned helplessness model of depression. Nature 470:535\u0026ndash;539\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetzger M, Bueno D, Lima LB (2017) The lateral habenula and the serotonergic system. Pharmacol Biochem Behav 162:22\u0026ndash;28\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao H, Zhang B-L, Yang S-J, Rusak B (2015) The role of lateral habenula\u0026ndash;dorsal raphe nucleus circuits in higher brain functions and psychiatric illness. Behav Brain Res 277:89\u0026ndash;98\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaker PM et al (2016) The lateral habenula circuitry: Reward processing and cognitive control. J Neurosci 36:11482\u0026ndash;11488\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLammel S et al (2012) Input-specific control of reward and aversion in the ventral tegmental area. Nature 491:212\u0026ndash;217\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKawai T, Yamada H, Sato N, Takada M, Matsumoto M (2015) Roles of the lateral habenula and anterior cingulate cortex in negative outcome monitoring and behavioral adjustment in nonhuman primates. Neuron 88:792\u0026ndash;804\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawson RP et al (2017) Disrupted habenula function in major depression. Mol Psychiatry 22:202\u0026ndash;208\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEly BA, Stern ER, Kim J-w, Gabbay V, Xu J (2019) Detailed mapping of human habenula resting-state functional connectivity. NeuroImage 200:621\u0026ndash;634\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Z, Wang S, Lee TMC, Zhang R (2023) Habenula functional connectivity variability increases with disease severity in individuals with major depression. J Affect Disord 333:216\u0026ndash;224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTorrisi S, Nord CL, Balderston NL, Roiser JP, Grillon C, Ernst M (2017) Resting state connectivity of the human habenula at ultra-high field. NeuroImage 147:872\u0026ndash;879\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEly BA, Xu J, Goodman WK, Lapidus KA, Gabbay V, Stern ER (2016) Resting-state functional connectivity of the human habenula in healthy individuals: Associations with subclinical depression. Hum Brain Mapp 37:2369\u0026ndash;2384\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBian B et al (2023) Recent advances in habenula imaging technology: A comprehensive review. J Magn Reson Imaging\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDumoulin SO, Fracasso A, van der Zwaag W, Siero JCW, Petridou N (2018) Ultra-high field MRI: Advancing systems neuroscience towards mesoscopic human brain function. NeuroImage 168:345\u0026ndash;357\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarrison OK, Guell X, Klein-Fl\u0026uuml;gge MC, Barry RL (2021) Structural and resting state functional connectivity beyond the cortex. NeuroImage 240:118379\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eViessmann O, Polimeni JR (2021) High-resolution fMRI at 7Tesla: challenges, promises and recent developments for individual-focused fMRI studies. Curr Opin Behav Sci 40:96\u0026ndash;104\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston KJ, Harrison L, Penny W (2003) Dynamic causal modelling. NeuroImage 19:1273\u0026ndash;1302\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston KJ et al (2019) Dynamic causal modelling revisited. NeuroImage 199:730\u0026ndash;744\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStephan KE, Penny WD, Moran RJ, den Ouden HEM, Daunizeau J, Friston KJ (2010) Ten simple rules for dynamic causal modeling. NeuroImage 49:3099\u0026ndash;3109\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteward T et al (2022) A thalamo-centric neural signature for restructuring negative self-beliefs. Mol Psychiatry 27:1611\u0026ndash;1617\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark GI, Egan SJ (2015) The Socratic method in Cognitive Behavioural Therapy: A narrative review. Cognit Ther Res 39:863\u0026ndash;879\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeck JS (2021) Cognitive behavior therapy: Basics and beyond, 3rd edn. The Guilford Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCooper M, Cohen-Tov\u0026eacute;e E, Todd G, Wells A, Tov\u0026eacute;e M (1997) The Eating Disorder Belief Questionnaire: Preliminary development. Behav Res Ther 35:381\u0026ndash;388\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEhring T, Zetsche U, Weidacker K, Wahl K, Sch\u0026ouml;nfeld S, Ehlers A (2011) The Perseverative Thinking Questionnaire (PTQ): Validation of a content-independent measure of repetitive negative thinking. J Behav Ther Exp Psychiatry 42:225\u0026ndash;232\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeidman P et al (2019) A guide to group effective connectivity analysis, Part 1: First level analysis with DCM for fMRI. NeuroImage 200:174\u0026ndash;190\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeidman P et al (2019) A guide to group effective connectivity analysis, Part 2: Second level analysis with PEB. NeuroImage 200:12\u0026ndash;25\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreaves MD, Novelli L, Razi A (2024) Structurally informed resting-state effective connectivity recapitulates cortical hierarchy. \u003cem\u003ebioRxiv\u003c/em\u003e, 2024.2004.2003.587831\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSokolov AA, Zeidman P, Erb M, Ryvlin P, Pavlova MA, Friston KJ (2019) Linking structural and effective brain connectivity: structurally informed Parametric Empirical Bayes (si-PEB). Brain Struct Funct 224:205\u0026ndash;217\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStephan KE, Tittgemeyer M, Kn\u0026ouml;sche TR, Moran RJ, Friston KJ (2009) Tractography-based priors for dynamic causal models. NeuroImage 47:1628\u0026ndash;1638\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston KJ et al (2016) Bayesian model reduction and empirical Bayes for group (DCM) studies. NeuroImage 128:413\u0026ndash;431\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatsumoto M, Hikosaka O (2009) Representation of negative motivational value in the primate lateral habenula. Nat Neurosci 12:77\u0026ndash;84\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteward T et al (2020) Dynamic neural interactions supporting the cognitive reappraisal of emotion. Cereb Cortex 00:1\u0026ndash;13\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeech R, Sharp DJ (2014) The role of the posterior cingulate cortex in cognition and disease. Brain 137:12\u0026ndash;32\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeech R, Kamourieh S, Beckmann CF, Sharp DJ (2011) Fractionating the Default Mode Network: Distinct contributions of the ventral and dorsal posterior cingulate cortex to cognitive control. J Neurosci 31:3217\u0026ndash;3224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFoster BL, Koslov SR, Aponik-Gremillion L, Monko ME, Hayden BY, Heilbronner SR (2023) A tripartite view of the posterior cingulate cortex. Nat Rev Neurosci 24:173\u0026ndash;189\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndrews-Hanna JR, Smallwood J, Spreng RN (2014) The default network and self-generated thought: Component processes, dynamic control, and clinical relevance. Ann N Y Acad Sci 1316:29\u0026ndash;52\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenon V (2023) 20 years of the default mode network: A review and synthesis. Neuron 111:2469\u0026ndash;2487\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavey CG, Pujol J, Harrison BJ (2016) Mapping the self in the brain's default mode network. NeuroImage 132:390\u0026ndash;397\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgathos J et al (2023) Differential engagement of the posterior cingulate cortex during cognitive restructuring of negative self- and social beliefs. Soc Cogn Affect Neurosci 18\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaker PM, Mizumori SJY (2017) Control of behavioral flexibility by the lateral habenula. Pharmacol Biochem Behav 162:62\u0026ndash;68\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrsini CA, Moorman DE, Young JW, Setlow B, Floresco SB (2015) Neural mechanisms regulating different forms of risk-related decision-making: Insights from animal models. Neurosci Biobehav Rev 58:147\u0026ndash;167\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaker PM, Raynor SA, Francis NT, Mizumori SJY (2017) Lateral habenula integration of proactive and retroactive information mediates behavioral flexibility. Neuroscience 345:89\u0026ndash;98\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStalnaker TA, Cooch NK, Schoenbaum G (2015) What the orbitofrontal cortex does not do. Nat Neurosci 18:620\u0026ndash;627\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlein-Fl\u0026uuml;gge MC, Bongioanni A, Rushworth MFS (2022) Medial and orbital frontal cortex in decision-making and flexible behavior. Neuron 110:2743\u0026ndash;2770\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiv Y (2019) Learning task-state representations. Nat Neurosci 22:1544\u0026ndash;1553\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudebeck PH, Rich EL (2018) Orbitofrontal cortex. Curr Biol 28:R1083\u0026ndash;R1088\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchuck Nicolas W, Cai Ming B, Wilson Robert C, Niv Y (2016) Human orbitofrontal cortex represents a cognitive map of state space. Neuron 91:1402\u0026ndash;1412\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBromberg-Martin ES, Hikosaka O (2011) Lateral habenula neurons signal errors in the prediction of reward information. Nat Neurosci 14:1209\u0026ndash;1216\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMizumori SJY, Baker PM (2017) The lateral habenula and adaptive behaviors. Trends Neurosci 40:481\u0026ndash;493\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathis V, Barbelivien A, Majchrzak M, Mathis C, Cassel J-C, Lecourtier L (2016) The lateral habenula as a relay of cortical information to process working memory. Cereb Cortex 27:5485\u0026ndash;5495\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhalighinejad N, Garrett N, Priestley L, Lockwood P, Rushworth MFS (2021) A habenula-insular circuit encodes the willingness to act. Nat Commun 12:6329\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNamboodiri VMK, Rodriguez-Romaguera J, Stuber GD (2016) The habenula. Curr Biol 26:R873\u0026ndash;R877\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSachella TE et al (2022) A novel role for the lateral habenula in fear learning. Neuropsychopharmacology 47:1210\u0026ndash;1219\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathis V, Cosquer B, Avallone M, Cassel J-C, Lecourtier L (2015) Excitatory transmission to the lateral habenula Is critical for encoding and retrieval of spatial memory. Neuropsychopharmacology 40:2843\u0026ndash;2851\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMathis V, Lecourtier L (2017) Role of the lateral habenula in memory through online processing of information. Pharmacol Biochem Behav 162:69\u0026ndash;78\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheehan DV et al (1998) The Mini-International Neuropsychiatric Interview (MINI): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 59:22\u0026ndash;33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmerican Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders: DSM-5, 5th edn. American Psychiatric Association\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInce S et al (2023) Subcortical contributions to salience network functioning during negative emotional processing. NeuroImage 270:119964\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOsmo F et al (2018) The Negative Core Beliefs Inventory: Development and psychometric properties. J Cogn Psychother 32:67\u0026ndash;84\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoeller S et al (2010) Multiband multislice GE-EPI at 7 tesla, with 16-fold acceleration using partial parallel imaging with application to high spatial and temporal whole-brain fMRI. Magn Reson Med 63:1144\u0026ndash;1153\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun H et al (2020) Extracting more for less: Multi-echo MP2RAGE for simultaneous T1-weighted imaging, T1 mapping, mapping, SWI, and QSM from a single acquisition. Magn Reson Med 83:1178\u0026ndash;1191\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarques JP, Kober T, Krueger G, van der Zwaag W, Van de Moortele P-F, Gruetter R (2010) MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field. NeuroImage 49:1271\u0026ndash;1281\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilke M (2012) An alternative approach towards assessing and accounting for individual motion in fMRI timeseries. NeuroImage 59:2062\u0026ndash;2072\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAshburner J (2007) A fast diffeomorphic image registration algorithm. NeuroImage 38:95\u0026ndash;113\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKasper L et al (2017) The PhysIO toolbox for modeling physiological noise in fMRI data. J Neurosci Methods 276:56\u0026ndash;72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReynaud O, Jorge J, Gruetter R, Marques JP, van der Zwaag W (2017) Influence of physiological noise on accelerated 2D and 3D resting state functional MRI data at 7 T. Magn Reson Med 78:888\u0026ndash;896\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlover GH, Li TQ, Ress D (2000) Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR. Magn Reson Med 44:162\u0026ndash;167\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBirn RM, Smith MA, Jones TB, Bandettini PA (2008) The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration. NeuroImage 40:644\u0026ndash;654\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang C, Cunningham JP, Glover GH (2009) Influence of heart rate on the BOLD signal: the cardiac response function. NeuroImage 44:857\u0026ndash;869\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBehzadi Y, Restom K, Liau J, Liu TT (2007) A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. NeuroImage 37:90\u0026ndash;101\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChakravarty MM et al (2013) Performing label-fusion-based segmentation using multiple automatically generated templates. Hum Brain Mapp 34:2635\u0026ndash;2654\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGermann J et al (2020) Fully Automated Habenula Segmentation Provides Robust and Reliable Volume Estimation Across Large Magnetic Resonance Imaging Datasets, Suggesting Intriguing Developmental Trajectories in Psychiatric Disease. Biol Psychiatry Cogn Neurosci Neuroimaging 5:923\u0026ndash;929\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhumada-Galleguillos P, Lemus CG, D\u0026iacute;az E, Osorio-Reich M, H\u0026auml;rtel S, Concha ML (2017) Directional asymmetry in the volume of the human habenula. Brain Struct Funct 222:1087\u0026ndash;1092\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRanft K, Dobrowolny H, Krell D, Bielau H, Bogerts B, Bernstein HG (2010) Evidence for structural abnormalities of the human habenular complex in affective disorders but not in schizophrenia. Psychol Med 40:557\u0026ndash;567\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieto-Castanon A, Whitfield-Gabrieli S (2022) \u003cem\u003eCONN functional connectivity toolbox: RRID SCR_009550 release 22\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlszowy W, Aston J, Rua C, Williams GB (2019) Accurate autocorrelation modeling substantially improves fMRI reliability. Nat Commun 10:1220\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston K (2009) Causal modelling and brain connectivity in functional magnetic resonance imaging. PLoS Biol 7:e1000033\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePenny WD, Stephan KE, Mechelli A, Friston KJ (2004) Comparing dynamic causal models. NeuroImage 22:1157\u0026ndash;1172\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKahan J, Foltynie T, Understanding DCM (2013) Ten simple rules for the clinician. NeuroImage 83:542\u0026ndash;549\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston K, Penny W (2011) Post hoc Bayesian model selection. NeuroImage 56:2089\u0026ndash;2099\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosa MJ, Friston K, Penny W (2012) Post-hoc selection of dynamic causal models. J Neurosci Methods 208:66\u0026ndash;78\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXia M, Wang J, He Y (2013) BrainNet Viewer: A network visualization tool for human brain connectomics. PLoS ONE 8:e68910\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5634827/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5634827/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSelf-related cognitions are integral to personal identity and psychological wellbeing. Persistent engagement with negative self-cognitions can precipitate mental ill health; whereas the ability to restructure them is protective. Here, we leverage ultra-high field 7T fMRI and dynamic causal modelling to characterise a negative self-cognition network centred on the habenula – a small midbrain region linked to the encoding of punishment and negative outcomes. We model habenula effective connectivity in a discovery sample of healthy young adults (\u003cem\u003en\u003c/em\u003e = 48) and in a replication cohort (\u003cem\u003en\u003c/em\u003e = 56) using a novel cognitive restructuring task during which participants repeated or restructured negative self-cognitions. The restructuring of negative self-cognitions elicits an excitatory effect from the habenula to the posterior orbitofrontal cortex that is reliably observed across both samples. Furthermore, we identify an excitatory effect of the habenula on the posterior cingulate cortex during both the repeating and restructuring of self-cognitions. Our study provides the first evidence in humans demonstrating the habenula’s contribution to processing self-cognitions. These findings yield novel insights into habenula’s function beyond processing external reward/punishment to include abstract internal experiences.\u003c/p\u003e","manuscriptTitle":"Habenula neural circuitry drives negative self-cognitions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-07 12:00:28","doi":"10.21203/rs.3.rs-5634827/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"d2b56d97-1ec1-42f6-ae7c-175130fb1ea8","owner":[],"postedDate":"January 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42419525,"name":"Biological sciences/Neuroscience/Cognitive neuroscience"},{"id":42419526,"name":"Biological sciences/Neuroscience/Computational neuroscience/Network models"},{"id":42419527,"name":"Health sciences/Health care/Medical imaging/Magnetic resonance imaging"},{"id":42419528,"name":"Biological sciences/Neuroscience/Neural circuits"},{"id":42419529,"name":"Biological sciences/Psychology/Human behaviour"}],"tags":[],"updatedAt":"2025-05-08T07:05:31+00:00","versionOfRecord":{"articleIdentity":"rs-5634827","link":"https://doi.org/10.1038/s41467-025-59611-7","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-05-07 04:00:00","publishedOnDateReadable":"May 7th, 2025"},"versionCreatedAt":"2025-01-07 12:00:28","video":"","vorDoi":"10.1038/s41467-025-59611-7","vorDoiUrl":"https://doi.org/10.1038/s41467-025-59611-7","workflowStages":[]},"version":"v1","identity":"rs-5634827","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5634827","identity":"rs-5634827","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}