Zusammenfassung: Die Biologische Psychologie und Neuropsychologie (BNP) untersucht die biologischen Grundlagen psychischer Prozesse – von molekularen Mechanismen bis hin zu endokrinen Regulationssystemen. Sie steht dabei in einer wechselseitigen Beziehung zur künstlichen Intelligenz (KI): BNP-Forschung inspirierte frühe KI-Modelle, während KI heute zunehmend die BNP beeinflusst. Dieser Beitrag beleuchtet KI aus Sicht der BNP entlang dreier Dimensionen: Grundlagen, Anwendungen und Grenzen. Als Analysewerkzeug ermöglicht NeuroKI/NeuroAI in der BNP-Forschung die Auswertung hochdimensionaler, multimodaler biologischer Datensätze sowie eine individualisiertere Diagnostik. Künstliche neuronale Netze dienen darüber hinaus als mechanistische Modelle neuronaler Informationsverarbeitung: Das Feld der NeuroAI vergleicht interne Repräsentationen von Gehirn und künstlichen neuronalen Netzen systematisch, um wechselseitig bessere Modelle und empirisch prüfbare Hypothesen zu entwickeln. Im Bereich der Anwendungen eröffnet KI als Stimulusgenerator, Interaktionspartner und Schnittstelle in körpernahen Technologien (z.B. Brain-Computer-Interfaces) neue experimentelle und therapeutische Möglichkeiten. In der BNP-Lehre kann KI als Coding-Assistent und individueller Tutor den Kompetenzerwerb beschleunigen. Gleichzeitig bestehen zentrale Grenzen und Risiken: mangelnde Transparenz (Black Box), Bias in Trainingsdaten, eingeschränkte Reproduzierbarkeit sowie ethische Fragen zu Autonomie, Verantwortlichkeit und Nachhaltigkeit. In der Lehre droht zudem ein Verlust wissenschaftlicher Kernkompetenzen. Die BNP kann umgekehrt zur menschenzentrierten KI-Entwicklung beitragen, indem biopsychologische Erkenntnisse zu kognitiven Grenzen, interindividuellen Unterschieden, Stressreaktivität und Embodiment in KI-Design und -Regulierung einfließen und allzu einfache Analogien zwischen biologischer und künstlicher Intelligenz korrigiert werden.

Abstract Recent adversity (RA) increases the risk of later mental health problems, but the mechanisms underlying this association remain largely unclear. Altered associative learning and specifically threat learning processes may constitute a potential pathway. Cross-sectional studies link RA to reduced threat-safety discrimination, a process crucial for adaptive functioning; however, longitudinal evidence allowing for more causal conclusions is lacking. In this prospective longitudinal study, 69 healthy participants completed fear acquisition, 24h-delayed extinction training and a reinstatement-test at two measurement time points six months apart (T0 and T1). We captured exposure to RA between T0 and T1 as well as autonomic and neural responses (SCRs and BOLD-fMRI), fear ratings, and salivary and hair cortisol at both assessments. During acquisition training, at the beginning of extinction (i.e., fear recall) and reinstatement-test at T1, RA-exposed individuals showed less autonomic threat-safety discrimination than unexposed individuals, mainly driven by blunted threat signal responding. These group differences predicted depression levels at the 1 and 1.5 year follow-ups and were mirrored by distinct activation in key fear-related brain regions, including striatal regions, thalamus, vmPFC, insula, and amygdala, whereas self-reported fear was unrelated to RA. In conclusion, across multiple outcomes, RA shapes threat learning and retrieval under conditions of (potential) imminent threat, with reduced autonomic threat discrimination prospectively predicting elevated depressive symptoms at 1- and 1.5-year follow-up. Overall, these findings implicate altered threat learning and retrieval processes as a potential mechanistic pathway through which RA becomes biologically embedded, potentially elevating psychopathological risk.

Patients with anxiety disorders overgeneralize conditioned defensive responses to cues, which have never been associated with threat and should therefore be considered safe. Such overgeneralization can be impairing calling for strategies to reduce it. Discrimination training may offer a simple way to enhance individuals' ability to distinguish between threat-predicting stimuli and similar but safe cues. In this study, we investigated the effects of discrimination training on fear generalization in 224 healthy participants who differed in their level of anxiety. Following a classical acquisition and a generalization phase, participants underwent discrimination training with either learning-relevant or learning-irrelevant stimuli. Additionally, half of the participants received feedback on their performance during training. We found significant reduction in generalization of all conditioned defensive responses (ratings and skin conductance), independent of the type of stimuli used during training. Feedback ameliorated the generalization of US-expectancy ratings, whereas the reduction of affective ratings and physiological responses was greater when no feedback was provided. Individuals with higher level of anxiety benefited more pronouncedly from the discrimination training. Discrimination training led to stronger reduction of the generalization of subjective arousal in individuals with higher level of anxiety. In summary, discrimination training is a simple yet effective method to reduce generalization of conditioned defensive responses, and anxious individuals showed the greatest benefit from this simple training. Such beneficial discrimination effects have important implications for treatments as such a training could be easily integrated into therapeutic interventions.

OBJECTIVES:Anxiety disorders involve abnormal fear learning and exaggerated generalization, proposed as potential risk factors. However, it is unclear whether these processes reflect stable vulnerability markers or symptoms of psychopathology, as most research relies on cross-sectional correlations and few longitudinal studies assess their predictive role in anxiety. Furthermore, little is known about the test-retest reliability of fear learning paradigms, despite their suggested function as trait-like markers for anxiety. METHODS:This longitudinal study reassesses initially healthy adults after 3-12 years, and children first examined at ages 8-12, then 12-16, and now at 16-22. Participants repeat the "screaming lady" paradigm, including fear acquisition and generalization. We integrate fear learning, psychometrics, HPA-axis activity, and (epi)genetics to assess the stability of fear generalization, its predictive value for later anxiety, and the link of fear learning to the HPA system. CONCLUSION:Our findings will clarify the long-term stability of fear learning paradigms and evaluate their value as markers of anxiety risk. In addition, a random forest model will explore multiple risk factors and their interaction in anxiety symptomatology. This may improve understanding of anxiety disorders and support early detection and prevention. TRIAL REGISTRATION:Retrospectively registered on AsPredicted.org (#238,927).

Affective experiences are inevitably accompanied by physiological changes. However, it is still a matter of intense debate whether events evoking similar affective experiences produce comparable physiological responses (fingerprint hypothesis) or variation is the norm within individuals (populations hypothesis). We reanalyzed data from two independent samples (N = 695; N = 64), using representational similarity analysis (RSA) to examine the trial-by-trial similarity patterns of subjective experience of valence and arousal and affect-related physiological measures (skin conductance [SCR] and startle blink responses) and their modulation by physiological intraindividual variability. Across physiological measures and tasks (passive picture viewing, passive sound listening, and imagery tasks), we observed that disregarding versus considering intraindividual variability when constructing psychophysiological response patterns yielded different associations with affective models, aligning with the fingerprint hypothesis and the populations hypotheses of affect, respectively. More importantly, follow-up analysis revealed that considering intraindividual variability yields a better representation of the SCR and startle patterns across individuals and tasks. Our results demonstrated that similar affective experiences are rather reflected by distinct physiological responses and emphasized the importance of considering intraindividual variability in future studies to better understand how physiological changes contribute to conscious affective experiences in humans.

Defensive responses to threat are critical for survival and often involve freezing-like behavior, which has been linked to action preparation. However, it remains unclear how this preparatory mechanism is influenced by unpredictable aversive stimuli that are difficult to discriminate or anticipate. In this study, we introduce a modified paradigm to examine how freezing-like behavior is affected by action preparation under temporal threat predictability, unpredictability, and safety. Using a multimodal approach in a large sample (n = 235, 148 female), we assessed postural sway and heart rate as markers of freezing-like behavior, alongside skin conductance levels, electromyographic startle responses, behavioral performance, and subjective ratings. Behaviorally, participants responded faster but less accurately under threat compared to safety, replicating previous findings. Physiologically, threat conditions were associated with both parasympathetically mediated postural freezing and sympathetically driven increases in skin conductance. Importantly, the temporal (un)predictability of threat did not modulate freezing-like behavior or skin conductance responses. Startle responses were generally inhibited during all conditions (threat and safety) relative to the inter-trial interval. These findings indicate that action preparation under threat involves concurrent activation of both parasympathetic (postural freezing) and sympathetic systems (skin conductance), consistent with animal models of freezing. The absence of modulation by temporal threat unpredictability may be indicative of a generalized defensive mode that prioritizes readiness over specificity.

Anxiety is heritable and exists on a continuum, with symptoms ranging from adaptive threat response to clinical disorder. Here we performed a genome-wide association meta-analysis of generalized anxiety symptom severity in 693,869 individuals of European ancestry from 14 cohorts. We identified 80 independent genome-wide significant variants within 74 loci, 39 of which were newly associated with anxiety. SNP-based heritability was 5.9% (posterior s.d. = 0.15%). Polygenic scores were significantly associated with anxiety symptom severity and disorder in European, African and South Asian ancestry samples (R2 = 1.2-2.9%). Significant genetic correlations (rg) were estimated with mental and physical health traits, including case-control anxiety, neuroticism and depression (rg = 0.71-0.85), irritable bowel syndrome (rg = 0.57), coronary artery disease, endometriosis and migraine (rg = 0.20-0.27). Gene-based and pathway analyses implicated synaptic and axonal processes, with enriched expression in the brain. These findings highlight the discovery power gained from analysing a quantitative trait rather than a case-control phenotype in anxiety genetics.

Background: By studying how individuals in an "at-risk" state of psychosis learn about threat and safety cues – specifically, how they develop and unlearn fear responses to neutral cues - we might better understand the mechanisms leading to heightened arousal and fear that are characteristic of acute psychotic episodes. Methods: At-risk individuals (N = 88; of which 28 fulfilled ultra-high-risk criteria on the Comprehensive Assessment of At-Risk Mental States interview and 60 scored above a predefined threshold in the Community assessment of Psychic Experiences questionnaire) and healthy controls (N = 44) underwent a standardized and validated differential fear conditioning paradigm including an acquisition, generalization, and extinction phase. The main outcomes of interest were the late positive potential, fear-potentiated startle, and self-reported ratings of valence, arousal, fear, and expectancy elicited by the conditioned stimuli (CS). Results: The at-risk group exhibited diminished fear learning, evident in significantly reduced differentiation between the CS+ vs. CS- in the valence ratings, compared to controls. Additionally, they demonstrated impaired fear extinction, evident in valence and arousal ratings, in which their CS differentiation showed a slower reduction than the controls. There were no group differences in late positive potential responses. Conclusion: At risk mental states appear to be associated with problems in distinguishing dangerous from safe stimuli and a diminished ability to adjust affective responses to conditioned stimuli based on new information, while the late-positive potential and fear-potentiated startle are unaltered. Early interventions could focus on recalibrating subjective emotional evaluations of fear-associated events.

Abstract Anxiety-related traits (ARTs) have been linked to altered fear learning, but previous studies have typically examined different experimental phases and response systems, limiting the comparability of findings and the accumulation of consistent evidence. Here, we comprehensively examined associations between ARTs and fear conditioning across acquisition, extinction and renewal and across subjective, physiological and neural response systems in a well-powered sample (N = 267) using a two-day differential conditioning paradigm. ARTs were operationalized as a composite of trait anxiety, neuroticism, and intolerance of uncertainty and conditioned responding was assessed using skin conductance responses, fear-potentiated startle, US expectancy ratings, fear ratings, and functional magnetic resonance imaging. Higher ARTs were consistently associated with elevated subjective fear and US expectancy to both threat and safety cues during extinction and renewal, without corresponding elevations in physiological responding. At the same time, ARTs were not associated with threat–safety discrimination in subjective or physiological measures across phases, while neural associations were limited to reduced dorsal anterior cingulate cortex discrimination during early renewal. These findings suggest that ARTs are characterized by a CS unspecific cognitive bias toward heightened threat expectancy and evaluation rather than altered associative fear learning, highlighting the importance of distinguishing conditioned discrimination from general levels of responding across response systems.

Neuroscience is crucial for understanding human behaviour. Yet, its resource-intensive methods contribute to the climate crisis. We call on neuroscientists to align their research with ecological sustainability goals across the research cycle and propose three key steps: replace unfocused data collection, reduce excessive emissions and refine imprecise methods.

Separation anxiety disorder (SEPAD) is characterized by pronounced fear or anxiety concerning separation from attachment figures. Despite its high lifetime prevalence, adult SEPAD often remains undetected due to a lack of diagnostic tools in multiple languages. The Adult Separation Anxiety Questionnaire (ASA-27) is a key instrument for assessing symptoms of SEPAD in adults. However, no validated German version is available. This study addressed the translation and validation of the ASA-27 in a German-speaking population to introduce the first German questionnaire assessing SEPAD. A consecutive forward and backward translation was conducted. Reliability and validity of the German ASA-27 against several established anxiety-related psychometric scores were assessed in a large sample of 1520 healthy participants. Results revealed robust internal consistency (Cronbach’s α = 0.87) and a factor structure explaining 49.7

We performed a genome-wide association meta-analysis of generalised anxiety symptom severity in 696,563 individuals of European ancestry from 14 cohorts. We identified 82 independent genome-wide significant variants within 76 loci, 41 of which were novel for anxiety. SNP-based heritability was 5.9% (SE = 0.19%). Polygenic scores were significantly associated with anxiety symptom severity and disorder in European, African, and South Asian ancestry samples (r2=1.2%-3.4%). Significant genetic correlations were estimated with numerous mental and physical health traits, including case-control anxiety, neuroticism and depression (rg=0.71-0.86), irritable bowel syndrome (rg=0.57), coronary artery disease, endometriosis, and migraine (rg=0.20-0.27). Gene-based and pathway analyses implicated synaptic and axonal processes, with enriched expression in the brain. These findings highlight the additional value of a dimensional approach in anxiety genetics.

The rapidly escalating climate crisis poses an existential threat to human wellbeing. Reducing anthropogenic greenhouse gas emissions must therefore become a primary goal of humanity. At the same time, advancing knowledge on human experience and behaviour through empirical research is likewise essential for wellbeing, but can incur substantial negative impact for the environment. Neuroscientific methods are particularly resource intensive and potentially harmful, from the carbon footprint of MRI scanners to the long-term impact of data centres keeping datasets permanently accessible for scientific reuse. This position paper addresses the resulting tension between scientific research, open science principles, and responsible scientific stewardship in times of the climate crisis. We discuss how sustainable open science practices can be implemented in neuroscience at each step of the research cycle following the ARIADNE framework. Specifically, we suggest to (1) re-place new data with open data, (2) re-fine methods to make them more sustainable, and (3) re-duce carbon emission of testing by precisely determining sample sizes and research protocols beforehand.

Background: The ability to adaptively transfer acquired fear to novel situations is fundamental for survival in ever-changing environments and may contribute to the emergence and persistence of anxiety disorders. Consequently, research has focused on the assessment of fear generalization profiles to predict individual differences in anxiety. However, substantial heterogeneity in the operationalization of generalization hampers comparisons across studies and poses a risk to the replicability of findings. Methods: To address these issues, we reviewed the literature to identify commonly used methods for characterizing perceptual fear generalization profiles. Then, we conducted simulation analyses to examine correlations between indices and probe their robustness against measurement noise. Finally, we used 2 large empirical datasets (N = 1175 and N = 256 healthy humans) to examine the reliability of these indices and their validity in predicting anxiety-related traits. Results: All identified indices were substantially correlated but highly sensitive to measurement noise, with only minimal differences between methods. Reliabilities were moderate for subjective ratings but poor for skin conductance responses. All indices of fear generalization were unrelated to anxiety-related traits. Conclusions: Overall, a more comprehensive discussion of conceptual and methodological issues is needed to enable informed decisions about how to reliably and validly estimate fear generalization and its relationship with anxiety-related traits or clinical symptoms.

Childhood adversity is a strong predictor of developing psychopathological conditions. Multiple theories on the mechanisms underlying this association have been suggested which, however, differ in the operationalization of ‘exposure.’ Altered (threat) learning mechanisms represent central mechanisms by which environmental inputs shape emotional and cognitive processes and ultimately behavior. 1402 healthy participants underwent a fear conditioning paradigm (acquisition training, generalization), while acquiring skin conductance responses (SCRs) and ratings (arousal, valence, and contingency). Childhood adversity was operationalized as (1) dichotomization, and following (2) the specificity model, (3) the cumulative risk model, and (4) the dimensional model. Individuals exposed to childhood adversity showed blunted physiological reactivity in SCRs, but not ratings, and reduced CS+/CS- discrimination during both phases, mainly driven by attenuated CS+ responding. The latter was evident across different operationalizations of ‘exposure’ following the different theories. None of the theories tested showed clear explanatory superiority. Notably, a remarkably different pattern of increased responding to the CS- is reported in the literature for anxiety patients, suggesting that individuals exposed to childhood adversity may represent a specific sub-sample. We highlight that theories linking childhood adversity to (vulnerability to) psychopathology need refinement.

The Spielberger State-Trait Anxiety Inventory (STAI) is the most cited measure of state and trait anxiety, and is routinely employed in a variety of research and clinical contexts. Here, we investigate the temporal stability as well as the convergent and discriminant validity of the German version of the STAI trait scale (STAI-T) across multiple time points in two independent samples (105 and 120 Caucasians). We observed temporal stabilities of .42-.67 for intervals between 20 and 41 months and from .81-.87 for intervals of five to 12 months, with decreasing stability as the time interval increased. Temporal stability estimates of the STAI-T were similar to those of related constructs. Additionally, examining the relationships within a nomological network support the recent conclusion that the STAI-T also shares substantial variance with questionnaires measuring negative emotionality such as depression, and hence does not measure anxiety specifically - despite its name. These results provide further psychometric information on what the STAI-T actually measures and to what extent STAI-T scores are expected to be stable across longer time intervals. This is of relevance for researchers aiming, for example, to use the STAI-T scale for predicting symptom trajectories and evaluating the effectiveness of therapeutic interventions.

Pavlovian fear conditioning is a fundamental process in both health and disease. We investigate its neural correlates and sources of variability using harmonized functional magnetic resonance imaging data from 2199 individuals across nine countries, including 1888 healthy individuals and 311 with anxiety-related or depressive disorders. Using mega-analysis and normative modeling, we show that fear conditioning consistently engages brain regions within the "central autonomic-interoceptive" or "salience" network. Several task variables strongly modulate activity in these regions, contributing to variability in neural responses. Additionally, brain activation patterns differ between healthy individuals and those with anxiety-related or depressive disorders, with distinct profiles characterizing specific disorders such as post-traumatic stress disorder and obsessive-compulsive disorder. While the neural correlates of fear conditioning are highly generalizable at the population level, variability arises from differences in task design and clinical status, highlighting the importance of methodological diversity in capturing fear learning mechanisms.

Adverse childhood experiences (ACEs) are a major risk factor for the development of multiple psychopathological conditions, but the mechanisms underlying this link are poorly understood. Associative learning encompasses key mechanisms through which individuals learn to link important environmental inputs to emotional and behavioral responses. ACEs may impact the normative maturation of associative learning processes, resulting in their enduring maladaptive expression manifesting in psychopathology. In this review, we lay out a systematic and methodological overview and integration of the available evidence of the proposed association between ACEs and threat and reward learning processes. We summarize results from a systematic literature search (following PRISMA guidelines) which yielded a total of 81 articles (threat: n=38, reward: n=43). Across the threat and reward learning fields, behaviorally, we observed a converging pattern of aberrant learning in individuals with a history of ACEs, independent of other sample characteristics, specific ACE types, and outcome measures. Specifically, blunted threat learning was reflected in reduced discrimination between threat and safety cues, primarily driven by diminished responding to conditioned threat cues. Furthermore, attenuated reward learning manifested in reduced accuracy and learning rate in tasks involving acquisition of reward contingencies. Importantly, this pattern emerged despite substantial heterogeneity in ACE assessment and operationalization across both fields. We conclude that blunted threat and reward learning may represent a mechanistic route by which ACEs may become physiologically and neurobiologically embedded and ultimately confer greater risk for psychopathology. In closing, we discuss potentially fruitful future directions for the research field, including methodological and ACE assessment considerations.