Existing evidence highlights sleep’s critical role in regulating cortisol stress recovery; the underlying neural pathways remain unclear. To address this gap, the current study aims to elucidate the neurobiological pathway linking objective sleep efficiency to cortisol stress recovery using functional magnetic resonance imaging (fMRI), with a focus on the functional connectivity (FC) between prefrontal cortex (PFC) and hippocampus.

Seventy-seven participants completed an acute stress task during a task-dependent and resting-state fMRI scanning. Salivary samples were collected and analyzed as an indicator of cortisol stress recovery. Objective sleep efficiency was measured the night before the fMRI scanning. Using Seed-based gPPI and resting-state FC analysis, we examined the mediating role of PFC-hippocampus FC in the association between objective sleep efficiency and cortisol stress recovery, both during the stress task and in the post-stress resting-state.

Objective sleep efficiency was significantly related to cortisol stress recovery but not with cortisol reactivity. Neurologically, higher sleep efficiency was linked to enhanced prefrontal activity and increased the left dlPFC-hippocampus FC during the acute stress task. Importantly, objective sleep efficiency promoted cortisol stress recovery by the weakened resting-state left dlPFC-hippocampus FC.

This study highlights the pivotal role of left dlPFC-hippocampus regulation underlying sleep’s effect on HPA axis recovery to acute stress. These results suggest a model whereby high objective sleep efficiency promotes adaptive stress recovery through dynamic reallocation of neural resources across acute stress process, characterized by task-dependent coupling and post-stress decoupling of frontal-hippocampal circuitry.

Emotion regulation relies on the interplay between prefrontal and limbic brain regions, with prefrontal regions implicated in the top-down modulation of the amygdala. In social anxiety disorder, disruptions in these networks have been reported, but most studies used undirected functional connectivity.

Dynamic causal modelling (DCM) was used to assess effective (i.e. directed) connectivity differences during emotion processing and regulation in individuals with social anxiety disorder compared with healthy controls.

A total of 102 participants (61 with social anxiety disorder, 41 healthy controls) performed a functional magnetic resonance imaging emotion regulation task under two conditions: viewing neutral/negative faces, and downregulating emotions using a self-chosen strategy. DCM was applied to model effective connectivity among the amygdala and key prefrontal regions. Connectivity patterns were characterised in healthy controls, and group comparisons tested how social anxiety disorder differed from this baseline model using parametric empirical Bayes. Leave-one-out cross-validation (LOOCV) evaluated whether connectivity differences predicted diagnostic group, symptom severity and emotion regulation difficulties.

In healthy controls, observation of negative faces was characterised by reciprocal influences between the amygdala and prefrontal cortex (PFC), including increased amygdala-to-ventromedial PFC (vmPFC) connectivity and inhibitory vmPFC-to-amygdala connectivity. During emotion regulation, healthy controls showed negative modulation from the amygdala to all prefrontal regions. Patients with social anxiety disorder did not differ from controls in amygdala–prefrontal connectivity; their alterations were confined to prefrontal circuits, with inhibitory connectivity from the pre-supplementary motor area (preSMA) to dorsolateral PFC during observation and bidirectional excitatory connectivity between the preSMA and vmPFC during regulation. LOOCV indicated that connectivity differences predicted diagnostic group.

The results support the idea that emotion processing and regulation influence connectivity between prefrontal areas and the amygdala in a complex, feedback-driven manner. Our findings suggest that aberrant emotion regulation in social anxiety disorder appears to be more closely linked to differences in intra-prefrontal circuits than deficits in amygdala–prefrontal connectivity.

Working memory (WM) impairment is a core cognitive deficit in schizophrenia, associated with dysfunction of large-scale brain networks, particularly the triple-network system comprising the default mode, frontoparietal, and salience networks. Given the role of environmental risks like childhood trauma (CT) in cognitive deficits, we investigated whether trauma relates to altered triple-network flexibility and WM in schizophrenia.

We enrolled 190 patients with schizophrenia (SZ) and 117 healthy controls (HCs). Among them, 162 SZ and 99 HCs underwent n-back task-based functional magnetic resonance imaging. We computed temporal variability (TV) in the triple-network connectivity, defining ΔTV as the change between 0-back and 2-back conditions. Subgroup comparisons of ΔTV were conducted within each group based on trauma status. Associations of ΔTV with WM performance and clinical symptoms were examined in SZ, followed by mediation analyses testing whether ΔTV mediates the relationship between trauma and WM.

Among HCs, individuals with childhood trauma showed reduced ΔTV across triple-network connections, whereas no such differences appeared in SZ. In SZ, greater ΔTV within the frontoparietal network (FPN) was correlated with lower positive symptom severity (r = −0.211, p-fdr = 0.046) and better n-back target accuracy (r = 0.303, p-fdr = 0.002). Furthermore, ΔTV within the FPN partially mediated the association between trauma and n-back accuracy.

Our findings highlight the central role of FPN flexibility in mediating childhood trauma’s effect on working memory in schizophrenia. This outlines a key pathway through which an early environmental risk (trauma) translates into cognitive and clinical manifestations in schizophrenia.

The quest for non-invasive and cost-effective biomarkers for mild cognitive impairment (MCI) and Alzheimer’s disease (AD) has led to growing interest in resting-state functional magnetic resonance imaging (MRI). This study examined associations between whole-brain functional connectivity measures and cognitive performance across a spectrum of cognitive aging.

A total of 108 older adults (mean age 74.1 ± 5.7 years), comprised of cognitively intact individuals, participants with amnestic MCI, and those with mild dementia due to probable AD, underwent high-resolution structural MRI and resting-state functional MRI scans and cognitive testing with the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Functional connectivity values were derived from a 17-network brain parcellation. Correlations were established between network connectivity values and RBANS Index scores.

Analyses revealed that lower RBANS Attention Index and Total Scale scores were significantly associated with increased connectivity between the ventral attention, central executive network, and limbic and default mode networks. Lower RBANS total scores were also associated with functional connectivity strength between the dorsal default mode networks and lateral frontoparietal regions of the central executive network, with increased connectivity observed across the dementia spectrum (Intact-MCI-AD).

These findings suggest that aberrant and potentially compensatory increases in functional connectivity may be linked to cognitive decline, supporting the utility of resting-state functional MRI as a promising biomarker for MCI and AD.

Repetitive negative thinking (RNT) and neuroticism are risk factors for internalizing psychopathology. However, their interaction has only been investigated at the self-report level, and studies elucidating their interrelationship at the neural level are lacking. We therefore investigated the interaction of trait RNT and neuroticism with respect to the dynamics of neural networks during negative self-referential processing.

A sample of 110 healthy subjects reported trait RNT and neuroticism, followed by an RNT induction paradigm during fMRI. Dynamic coactivation pattern (CAP) analysis was used to identify a set of recurring coactivation patterns and to quantify their persistence and count rates. Next, the effects of trait RNT, neuroticism, and their interaction on brain dynamics were tested using regression models.

Negative interactions between RNT and neuroticism were found for persistence and counts of the canonical default mode network (DMN) as well as salience network (SAL) CAP. Simple slope analysis revealed that subjects scoring high on neuroticism exhibited a negative association between trait RNT and DMN as well as canonical SAL dynamics. Furthermore, trait RNT was positively associated with persistence and count rates of a hybrid FPN+DMN coactivation state.

Our results suggest that individuals with high neuroticism who spend more time in SAL and DMN CAPs may be less vulnerable to RNT, potentially reflecting more adaptive network configurations. Furthermore, less segregated CAPs, evident by the concurrent activation of functionally antagonistic networks (FPN+DMN), emerge more often in individuals prone to RNT, likely reflecting disrupted network interactions.

Adolescence is a sensitive period for social and neural development. Empathic growth during adolescence has been linked to improved prosocial behavior in adulthood. This study examined how adolescent empathy relates to adulthood neural responses to rejection.

Participants (N = 77; 42 females, 52% White) were drawn from a demographically diverse community sample and assessed annually from ages 13 to 21. Each year, participants’ empathic support provision toward a close friend was evaluated during an observationally coded support task. At approximately age 24, participants completed the Cyberball social exclusion paradigm while undergoing functional magnetic resonance imaging (fMRI).

Whole-brain exploratory analyses revealed that greater empathic support provision during adolescence was associated with reduced activation in the subgenual anterior cingulate cortex (sACC) during social exclusion in early adulthood (Cohen’s d = 0.12), suggesting a contribution of empathy provision to rejection-related neural responses later in life. The effect was not driven by felt distress during social exclusion, indicating that adolescent empathic support provision is potentially associated with neural responses to social exclusion independent of subjective distress.

These findings underscore the long-term links of empathy to adult social processes and may inform interventions aimed at enhancing interpersonal functioning and resilience.

Functional magnetic resonance imaging (fMRI) has revealed inconsistent neural activity patterns in major depressive disorder (MDD) across cognitive and affective domains, and this study used an activation likelihood estimation (ALE) meta-analysis to examine brain function abnormalities in working memory, reward processing, and emotion processing.

A systematic search was conducted in PubMed, Embase, Web of Science, ScienceDirect, and CNKI for fMRI studies comparing MDD patients with healthy controls (HCs), including data up to 3 December 2024. ALE meta-analysis was performed to examine activation patterns. Jackknife sensitivity analysis, risk of bias, and Newcastle–Ottawa scale were used to assess robustness and publication bias. Meta-regression analyses were conducted to explore the impact of covariates on the results.

Sixty-nine studies (2,073 MDD individuals and 2,009 HCs) were included. MDD individuals showed hyperactivation in the bilateral parahippocampal gyrus, subcallosal gyrus, lentiform nucleus, left claustrum, insula, and anterior cingulate cortex, alongside hypoactivation in the right lentiform nucleus, parahippocampal gyrus, fusiform gyrus, and other regions. Domain-specific analyses revealed working memory-related hyperactivation in the right middle and superior frontal gyrus, reward-related hyperactivation in the bilateral lentiform nucleus, right claustrum, and left caudate, and emotion-related hyperactivation in the bilateral parahippocampal gyrus, bilateral lentiform nucleus, right subcallosal gyrus, right anterior cingulate cortex, and left claustrum. Jackknife sensitivity analysis confirmed robustness, with no significant publication bias or covariate impact.

Aberrant activation in the lentiform and caudate nuclei across reward and emotion tasks suggests striatal dysfunction plays a key role in emotion-motivation interplay, highlighting the striatum as a potential target for future therapies.

Test whether a dissonance-based transdiagnostic eating disorder treatment, body project treatment (BPT), produces greater reduction in brain reward region response to the thin ideal and behaviors used to pursue this ideal and eating disorder symptoms, and higher abstinence from eating disorder behaviors and remittance from eating disorder diagnoses than a matched transdiagnostic interpersonal psychotherapy (IPT).

Women with various eating disorders (N = 83) were randomized to 8-week group-implemented BPT or IPT and completed functional magnetic resonance imaging (fMRI) at pretest and posttest, and surveys and masked diagnostic interviews at pretest, posttest, and 6-month follow-up.

BPT versus IPT participants showed significantly greater reductions in mid cingulate cortex response to thin models, anterior cingulate cortex response to eating disorder behavior words, eating disorder symptoms (d = 0.54), and body dissatisfaction (d = 0.57), and marginally greater reductions in psychosocial impairment (d = 0.39) at posttest, as well as significantly greater reductions in body dissatisfaction (d = 0.68) and psychosocial impairment (d = 0.63), and marginally greater reductions in eating disorder symptoms (d = 0.53) at 6-month follow-up. At posttest, BPT versus IPT participants showed significantly greater abstinence from binge eating and purging (48% versus 23%, respectively) but did not differ on remittance from eating disorder diagnoses (52% versus 44%, respectively).

Results provide further evidence of target engagement for BPT and suggest that it is more effective than IPT in treating a range of eating disorders.

Conduct disorder and childhood head injuries frequently co-occur and are linked to a higher risk of later delinquency. While both are known to disrupt reward-related neural circuits, this study investigated whether their combined presence leads to a unique disruption in these pathways, potentially accounting for the increased risk of delinquency.

Using neuroimaging data from the baseline (age 9–10) assessment from the Adolescent Brain and Cognitive Development (ABCD) study, four groups were compared: children with conduct disorder (CD, n = 588), a mild traumatic brain injury (mTBI, n = 1,216), both (mTBI+CD, n = 252), and typically developing controls (TD, n = 705). Neural activation in eight regions of interest (amygdala, hippocampus, nucleus accumbens, caudal anterior cingulate cortex, rostral anterior cingulate cortex, medial orbitofrontal cortex, thalamus, and insula) during reward anticipation and receipt were assessed during the monetary incentive delay task.

After controlling for several covariates, including sex, ADHD, and internalizing problems, the mTBI+CD group displayed greater left amygdala and hippocampal activation during reward receipt compared to all other groups. While they displayed increased activation in the right hippocampus and thalamus compared to TD controls and the right hippocampus compared to the mTBI group, they did not differ from the CD group. No group differences emerged during reward anticipation.

Increased left amygdala and hippocampus activation in children with conduct disorder and a history of mild traumatic brain injury may reflect robust encoding of emotionally charged reward experiences, potentially reinforcing memory-guided, reward-seeking behaviors.

Depression is characterized by divergent changes in positive and negative affect. Emerging roles of inflammation in depression portend avenues for novel immunomodulator-based monotherapy, targeting mechanistically distinct symptoms such as anhedonia and pessimism.

To investigate links between these divergent affective components and inflammation, we used a probabilistic reinforcement-learning fMRI paradigm, testing for evidence of hyposensitivity to reward, and hypersensitivity to punishment in low-inflammation depression cases (loCRP depression; CRP ≤ 3 mg/L; N = 48), high-inflammation depression cases (hiCRP depression; CRP > 3 mg/L; N = 31), and healthy controls (HC; CRP ≤ 3 mg/L; N = 45). We aimed to (i) determine whether depression cases with high and low inflammation showed aberrant neural activation to monetary gains and losses compared to controls, and (ii) examine if these alterations correlated with a continuous measure of C-reactive protein (CRP) in depression, as well as indices of anhedonia and pessimism derived from behavioral instruments in depression.

Voxel-wise activation was observed in key brain regions sensitive to monetary reward (ventromedial prefrontal cortex, vmPFC; nucleus accumbens, NAc) and punishment (insula) outcomes across all three groups. However, there was no significant difference in activation between groups. Within depression cases, increasing CRP scaled negatively with activation in the right vmPFC and left NAc but not insula cortex. However, there was no significant association between regional activation and severity of anhedonia or pessimism.

Our results support the previously reported association between CRP and striatal reward reactivity in depression but do not extend this to processing of negatively valenced information.

Anxiety is a persistent trait that disrupts functioning and increases the risk of severe consequences, while reward processing has garnered attention in anxiety research. Here, we report a critical concern in reward processing among individuals with anxiety: although anxious individuals may show similar reward processing abilities as non-anxious individuals in typical environments, they are more vulnerable to disruptions in positive emotions caused by frustrative non-reward, leading to maladaptive reward processing patterns.

The functional magnetic resonance imaging (fMRI) was used in this study. A total of 66 participants were recruited for the experiment, with 33 in the high anxiety (HA) group and 33 in the low anxiety (LA) group. The simulation of frustrative non-reward was conducted during fMRI scanning.

Under the low frustration condition, the HA group exhibited task accuracy comparable to the LA group and showed greater activation in visual processing regions (inferior occipital gyrus, superior occipital gyrus, angular gyrus) and cognitive control areas (precuneus, precentral gyrus) during attentional reorienting following frustration. However, in the high frustration condition, the HA group displayed significantly lower accuracy, with maladaptive information processing patterns observed in several brain regions associated with the cognitive-emotional control system (cuneus-precuneus, anterior cingulate cortex, precentral gyrus, inferior frontal gyrus, superior frontal gyrus, orbitofrontal cortex, and amygdala).

This demonstration of two contrasting processing patterns deepens the current understanding of reward processing in anxiety. It also holds significance for a broader understanding of the risk factors in cognitive processing among individuals with anxiety.

Repetitive negative thinking (RNT) in major depressive disorder (MDD) involves a persistent focus on negative self-related experiences. Resting-state fMRI shows that the functional connectivity (FC) between the anterior insula and the superior temporal sulcus is associated with RNT intensity. This study examines how insular FC patterns differ between resting state and RNT induction in MDD and healthy control (HC) participants.

Forty-one individuals with MDD and 28 HCs (total n = 69) underwent resting-state and RNT-induction fMRI scans. Seed-to-whole brain analysis using insular subregions as seeds was performed.

No diagnosis-by-run interaction effects were observed across insular subregions. MDD participants showed greater FC between the bilateral anterior, middle, and posterior insular regions and the cerebellum (z = 4.31–6.15). During RNT induction, both MDD and HC participants demonstrated increased FC between bilateral anterior/middle insula and prefrontal cortices, parietal lobes, posterior cingulate cortex (PCC), and medial temporal gyrus, encompassing the STS (z = 4.47–8.31). In exploratory correlation analyses, higher trait RNT was associated with increased FC between the right dorsal anterior/middle insula and the PCC, middle temporal gyrus, and orbital frontal gyrus in MDD participants (z = 4.31–6.15). Greater state RNT was linked to increased FC in similar insular regions, as well as the bilateral angular gyrus and right middle temporal gyrus (z = 4.47–8.31).

Hyperconnectivity in insula subregions during active rumination, especially involving the default mode network and salience network, supports theories of heightened self-focused and negative emotional processing in depression. These findings emphasize the neural basis of RNT when actively elicited in MDD.

Stress could increase delay discounting in subjects with bulimia nervosa and alcohol use disorder (AUD), meaning that the short-term benefits of coping through eating or drinking outweigh the long-term negative consequences. Therefore, this study explores differences in delay discounting between patients and healthy controls, the impact of stress on food and alcohol delay discounting and associated changes in brain activity.

A total of 102 female participants (AUD, 27; bulimia nervosa, 25; healthy controls, 50) underwent repeated functional magnetic resonance imaging scanning. Initially, all participants performed a monetary delay discounting task (DDT), followed by a food or alcohol DDT before and after stress induction. Specifically, those with bulimia nervosa completed a food DDT, those with AUD completed an alcohol DDT and healthy controls were randomly allocated to one or either DDT.

Participants with AUD, but not healthy controls, displayed a higher discounting of alcohol after stress. Healthy controls, but not those with bulimia nervosa, had nominally higher discounting rates of food following stress, although not significant following multiple testing correction. Participants with AUD displayed a lower activity of the right supplementary motor area while discounting alcohol after stress. Healthy controls showed a lower activity of the frontal cortex and a higher activity of the motor cortex while discounting food after stress, while those with bulimia nervosa displayed a higher activity of the occipital cortex.

The results suggest that, in subjects with AUD, stress induces neurobiological changes that cause them to prefer more immediately available alcohol. However, the results observed in participants with bulimia nervosa suggest a more complex relation between stress and food.

Maternal depressive symptoms can influence brain development in offspring, prenatally through intrauterine programming, and postnatally through caregiving related mother–child interaction.

The participants were 5-year-old mother–child dyads from the FinnBrain Birth Cohort Study (N = 68; 28 boys, 40 girls). Maternal depressive symptoms were assessed with the Edinburgh Postnatal Depression Scale (EPDS) at gestational week 24, 3 months, 6 months, and 12 months postnatal. Children’s brain imaging data were acquired with task-free functional magnetic resonance imaging (fMRI) at the age of 5 years in 7-min scans while watching the Inscapes movie. The derived brain metrics included whole-brain regional homogeneity (ReHo) and seed-based connectivity maps of the bilateral amygdalae.

We found that maternal depressive symptoms were positively associated with ReHo values of the left amygdala. The association was highly localized and strongest with the maternal depressive symptoms at 3 months postnatal. Seed-based connectivity analysis did not reveal associations between distal connectivity of the left amygdala region and maternal depressive symptoms.

These results suggest that maternal depressive symptoms soon after birth may influence offspring’s neurodevelopment in the local functional coherence in the left amygdala. They underline the potential relevance of postnatal maternal distress exposure on neurodevelopment that has received much less attention than prenatal exposures. These results offer a possible thus far understudied pathway of intergenerational effects of perinatal depression that should be further explored in future studies.