Amygdala: Threat, Salience, Learning, and Emotion
Author: Ukrainian Psychological Hub · Published: October 12, 2026 · Editorial Policy
The amygdala is a pair of interconnected groups of nuclei in the medial temporal lobes that helps the brain learn what matters, assign significance to events, guide attention, update associations between cues and outcomes, and organize bodily and behavioral responses. Threat learning is one of its most thoroughly established functions, but the same broader system also contributes to reward, social perception, emotionally significant memory, and flexible action. It is best understood through its component circuits and connections rather than as a single switch that produces fear. This account is supported by converging anatomy, human lesion work, conditioning experiments, and reviews of network-level emotion processing Pessoa and Adolphs, 2010, Sander et al., 2003.
What is the amygdala?
The amygdala is an anatomically defined complex of gray-matter nuclei deep within each cerebral hemisphere, toward the front of the medial temporal lobe. Its name comes from the Greek word for almond, referring to its approximate shape. The left and right amygdalae contain interacting neuronal populations with partly different connectivity and specialized contributions. They are components of distributed brain systems, linked to sensory cortices, hippocampus, prefrontal and cingulate regions, striatum, hypothalamus, and brainstem.
A useful functional definition is this: the amygdala contributes to learning and using the significance of sensory and internal events, particularly when those events predict biologically important consequences. Significance may concern danger, opportunity, uncertainty, motivational value, or socially informative signals. Its computations help influence which information is attended to, what associations are learned, what is remembered, and which response possibilities become more likely. This description captures the broader relevance and valuation literature while allowing distinct amygdala nuclei to perform different operations Pessoa, 2010.
It is also essential to distinguish a brain structure from the mental experiences associated with it. A learned defensive reaction, a self-reported feeling of fear, vigilance toward uncertainty, and the recognition of fear in another person's face are different outcomes. They may involve overlapping but nonidentical circuitry, and no one response is a substitute measurement for all the others.
Where is the amygdala located?
Each amygdala sits in the medial temporal lobe near the anterior hippocampus, within a larger landscape of temporal-lobe structures. The hippocampus is strongly involved in contextual, episodic, and spatial memory; the amygdala is particularly relevant to the significance, associative value, and behavioral consequences of events. Their close anatomical and functional interactions help explain why an emotionally consequential encounter can be both memorable and connected to a place or context.
The amygdala is commonly discussed as part of the limbic system. That label has historical value but does not identify a single anatomically uniform, functionally self-contained emotion apparatus. Modern research describes specific nuclei, pathways, and distributed networks, making the precise connections more informative than the broad limbic label.
On MRI, the amygdala can be anatomically delineated, though measurement of its small subregions depends on image resolution and segmentation methods. A structural scan indicates anatomy; it does not by itself measure whether a person is afraid, unusually anxious, or capable of regulating emotion. Similarly, an image showing a larger or smaller amygdala cannot be interpreted as a straightforward personality or clinical test.
The main nuclei and pathways: what is inside the amygdala?
Basolateral amygdala: association, sensory information, and value
The basolateral complex includes the lateral, basal, and accessory basal nuclei, although terminology and fine anatomical subdivisions vary across species and atlases. Broadly, these regions receive processed sensory and contextual input and contribute to learning relationships between events, stimuli, and valued or aversive outcomes. The basolateral complex also exchanges information with hippocampal and frontal cortical systems, supporting predictions and decisions that depend on context.
In classical conditioning, a previously neutral cue may acquire significance after it predicts an unpleasant event. Changes in activity and synaptic connections within relevant amygdala circuits contribute to that learning. Comparable circuit principles operate when a cue predicts something desirable. The emphasis is on learning relations between events rather than on storing one fixed emotional label for every stimulus Baxter and Murray, 2002.
Central amygdala: coordinating output and learning
The central nucleus has extensive connections with the hypothalamus and brainstem systems involved in physiological and behavioral responses. It contributes to changes in vigilance, autonomic activity, defensive action, and the control of other motivated behaviors. A traditional circuit diagram pictured it as an output relay after learning occurred elsewhere. Modern research shows that populations in the central amygdala can themselves participate in learning and plasticity.
Different central-amygdala cell populations are not interchangeable. A 2025 study recorded calcium signals from GABAergic central-amygdala neurons in mice during food and shock conditioning, showing heterogeneous responses associated with positive and negative outcomes. The authors' distinction between valence and salience coding remains partly debated in the accompanying peer review. These observations are valuable cellular evidence in mice, not proof that the same signal categories can be read directly from an individual human fMRI scan Kong et al., 2025.
Intercalated cells and other subdivisions
Clusters of inhibitory intercalated neurons between major amygdala nuclei can regulate information transfer and influence the balance among behavioral responses. Their connectivity, neuromodulation, and experience-dependent plasticity are active areas of investigation. A recent review emphasizes how these clusters coordinate interactions within the amygdala and with wider brain networks rather than simply turning fear off Aksoy-Aksel et al., 2025.
Medial and cortical or superficial amygdala divisions also contribute to processing social, olfactory, and biologically meaningful information, with important cross-species differences. Human research often has less cellular specificity than rodent experiments. Consequently, a claim about a named mouse cell type or optogenetically identified circuit should retain its species and experimental context when incorporated into a human-focused explanation.
What does the amygdala actually do?
1. Learning about threat and safety
A well-established role of amygdala circuitry is to support rapid learning when a cue predicts an aversive event. In Pavlovian threat-conditioning experiments, researchers may pair a visual image or sound with an unpleasant but controlled outcome. They then measure changes in skin conductance, startle, attention, or self-report when the predictive cue appears. These measures reveal distinct aspects of learned responding; they do not each establish the same subjective experience.
Human lesion studies offer unusually valuable causal evidence. In a study of four individuals with highly selective bilateral basolateral amygdala damage, acquisition of conditioned fear-potentiated startle was impaired, while declarative knowledge of the cue-outcome relationship could remain intact. This dissociation shows that knowing what predicts a threat and expressing a rapid conditioned defensive reflex can depend on different mechanisms Klumpers et al., 2015.
The response to a cue also depends on timing, context, previous learning, and uncertainty. A cue predicting possible harm in one setting may be safe in another. This is why hippocampal contextual signals and prefrontal mechanisms are relevant to amygdala-dependent learning. Learning is dynamic: it can update after circumstances change.
2. Responding to salience and relevance
Salience refers broadly to the importance or priority a stimulus acquires for attention and behavior. Some salient stimuli are threatening; others are rewarding, unexpected, novel, or socially meaningful. The relevance-detection account proposes that the amygdala helps assign priority to information with consequences for the organism's needs or goals Sander et al., 2003.
There are several concepts that should remain distinct. Valence concerns whether something is experienced or learned as favorable or unfavorable. Arousal refers to a dimension of activation or intensity. Salience concerns the prominence or behavioral importance of information. Threat probability concerns the estimated likelihood of danger. These variables can covary experimentally, making it difficult to determine which variable a particular regional signal represents.
The salience interpretation is therefore a useful model rather than an assertion that every amygdala neuron computes one universal salience quantity. Evidence points to multiple neuronal populations and network operations. The amygdala also interacts with the anterior insula and cingulate systems often discussed in relation to the larger salience network; its local anatomical function and the network-level concept are different levels of description.
3. Emotional learning and memory
Emotionally arousing events are often remembered differently from mundane ones. Evidence from animals and humans implicates the amygdala, particularly basolateral systems, in modulating the consolidation of memory through interactions with other structures. In a classic synthesis, McGaugh reviewed findings suggesting that stress-related hormonal and neuromodulatory processes influence the amygdala and its communication with memory circuits McGaugh, 2004.
The amygdala does not independently archive all the factual and spatial details of a life event. Hippocampal and cortical networks are important for aspects of episodic memory; amygdala-mediated modulation may affect how persistently significant aspects of an experience are retained. Nor does high emotion guarantee accurate recall. A vivid memory can be selective, incomplete, or mistaken, and emotional influence on memory varies with the task, timing, intensity, and outcome measured.
Acquisition, consolidation, retrieval, reconsolidation, generalization, and extinction are separate processes. A finding about any one stage must be described at that stage. Translating an effect on laboratory cue conditioning into a claim about complex autobiographical trauma memories requires additional evidence.
4. Reward, motivation, and positive significance
The amygdala participates in appetitive as well as aversive learning. Studies of animal behavior, neuronal responses, and human decision making indicate that amygdala circuits help represent the learned value and identity of anticipated outcomes, especially when choices depend on updating associations. Baxter and Murray's review placed stimulus-reward learning alongside the more familiar literature on fear conditioning Baxter and Murray, 2002.
An instructive contemporary example comes from a 2024 Nature Neuroscience study of dopamine projections to the basolateral amygdala. In a rodent model, these projections contributed to encoding memories that distinguished specific anticipated rewards, advancing an account of identity-specific prediction rather than a simplistic equation of dopamine with pleasure Sias et al., 2024.
This work illustrates why neurotransmitter language needs precision. Dopamine is involved in several forms of learning, motivation, and action; noradrenergic, cholinergic, glutamatergic, and GABAergic systems also modulate amygdala function. Individual transmitter actions vary across receptors, cellular populations, circuits, and timescales. The phrase "fear chemical" or "happiness chemical" cannot explain the amygdala. For a broader explanation of chemical signaling, see Neurotransmitters: How Chemical Messages Affect the Brain.
5. Attention to faces and social signals
Amygdala involvement in social perception is especially well illustrated by observations of the rare patient SM, who has extensive bilateral amygdala damage. In an influential experiment, her difficulties recognizing fear in faces were related to reduced spontaneous attention to the eye region. Directing her to look at the eyes improved performance Adolphs et al., 2005.
The result provides a mechanism for a particular impairment: altered selection of socially diagnostic information. It does not imply that a brain region contains an isolated mental faculty for reading emotions or that people who avoid eye contact have amygdala damage. Face interpretation involves distributed perceptual and contextual processes, and gaze behavior differs substantially across individuals and settings.
Human and animal research also examines how the amygdala contributes to processing social approach, avoidance, affiliation, ambiguity, and the value of social information. These findings are heterogeneous. Social behavior cannot be reduced to a stable quantity of "amygdala activation" or a single interpretation of an emotional-face experiment.
Does the amygdala create fear?
Fear is often used in research to refer to different things: a conscious feeling, a learned defensive response, a behavioral state, or a physiological pattern. Clarifying the intended meaning changes what the evidence supports. The amygdala plays a major role in particular threat-processing and defensive-learning operations, but the complete conscious experience of fear emerges within broader brain-body processes.
A striking pair of lesion findings demonstrates the distinction. In one study, patient SM exhibited severely diminished fear in response to several external threats, including animals and evocative experiences Feinstein et al., 2011. In a later study, three individuals with bilateral amygdala damage nevertheless experienced intense fear and panic when exposed experimentally to elevated carbon dioxide Feinstein et al., 2013. These small, exceptional samples do not settle every theory, but they directly challenge the claim that an intact amygdala is universally required for any form of fear or panic.
LeDoux and Pine proposed a two-system framework distinguishing circuits that organize defensive reactions from higher-order systems contributing to conscious fear and anxiety LeDoux and Pine, 2016. This is an influential theoretical model, not an uncontested final account of consciousness. Other distributed and constructionist approaches emphasize patterns of brain-body interaction and conceptual processing. The shared methodological lesson is to measure subjective experience separately from freezing, startle, autonomic arousal, and regional activity.
How does threat learning work?
In a standard learning experiment, an initially neutral cue becomes predictive because it is repeatedly paired with an unpleasant outcome. The conditioned response may strengthen as the participant learns that relationship. Neural plasticity within basolateral and central amygdala circuits, together with interactions among sensory, prefrontal, hippocampal, and brainstem systems, contributes to the acquisition and expression of responding.
During extinction, repeated presentation of the predictive cue without the adverse outcome often reduces the conditioned response. Extinction is generally modeled as new learning that alters the expression of previous associations, rather than as the deletion of an original memory trace. The older response can return when context changes, time passes, or certain events occur. A 2024 systematic review and meta-analysis of human fear-conditioning renewal experiments found that experimental design and context strongly influence the return of responding Wang et al., 2024.
The prefrontal cortex and hippocampus can contribute to adjusting threat responses in relation to goals and context. These interactions do not constitute a simple contest in which the rational cortex turns off an emotional amygdala. Multiple subregions have different and sometimes opposing effects, and the relevant direction depends on the task, timing, and stage of learning.
In March 2026, Meijer and colleagues reported a human experiment using transcranial ultrasound stimulation directed at the amygdala during cue-based conditioning. In 25 participants in the amygdala-targeted experiment, stimulation slowed early acquisition and was associated with more rapid subsequent extinction relative to sham, while a separate 25-person hippocampal targeting experiment served as a comparison. The investigators measured skin-conductance responses and fit learning models. This provides intervention-based evidence for a causal contribution to aspects of associative threat learning, while the sample sizes, target selectivity, replication needs, and experimental paradigm limit clinical generalization Meijer et al., 2026.
Amygdala connections with other brain regions
Hippocampus and context
The hippocampus participates in representing contexts and relationships among events. Interactions with the amygdala allow cues and environments to acquire and update significance. The response to an ambiguous sound can depend on whether it occurs in an expected, familiar setting or one associated with danger. Hippocampal processing does not merely provide a background image to which the amygdala mechanically assigns emotion; context and value are constructed through recurrent interactions.
Prefrontal cortex and behavioral flexibility
Prefrontal subdivisions contribute to task goals, valuation, uncertainty, reappraisal, and flexible control. Connectivity with the amygdala varies among pathways and cognitive states. Human connectivity measurements can support models of interaction, while demonstrating a directional causal mechanism generally requires more specific interventions or converging evidence. Statements that the prefrontal cortex simply "overrides" the amygdala erase this complexity.
Insula, body-state representations, and network salience
The insula is involved in processing internal bodily information and integrates with other regions implicated in attention and feeling states. Together with cortical and subcortical systems, it helps connect internal signals and external circumstances. The carbon-dioxide panic findings are one reason to consider interoceptive pathways when theorizing about fear: awareness of bodily disturbance can contribute to aversive experiences without following the same pathway as every external threat cue.
Hypothalamus and brainstem
Amygdala connections with the hypothalamus and brainstem participate in organizing autonomic, endocrine, attentional, and action-related changes. Heart rate and skin conductance reflect peripheral physiological processes under influence from many systems. They are informative measurements of arousal and conditioned responding, but neither alone identifies a specific emotion or its unique neural origin.
Is the left amygdala different from the right?
Researchers have investigated differences in activation, connectivity, temporal dynamics, and certain social or perceptual tasks between the two hemispheres. Some findings suggest lateralization in specific experimental conditions. Results depend on imaging methods, participant characteristics, tasks, and statistical choices, and there is no established everyday rule equating the left amygdala with one emotion and the right with another.
For readers interpreting a scan, lateralized activation must be linked to a specified contrast and a defined measure. It does not establish that one hemisphere is "the emotional side" of the person. Bilateral connections and coordination are typical features of brain function. For a fuller account of hemispheric specialization, see Left Brain vs. Right Brain: Hemispheres, Specialization, and Myths.
How scientists study the amygdala
Structural MRI
Structural magnetic resonance imaging uses magnetic fields and radiofrequency pulses to characterize tissue anatomy and contrast. It can help locate the amygdala, estimate volume, and detect some structural abnormalities. Size depends on segmentation protocols, age, individual variation, and other factors. A structural measurement cannot directly measure subjective fear, predict how someone will behave in a particular situation, or diagnose anxiety by itself.
Functional MRI
Functional MRI commonly measures a blood-oxygen-level-dependent (BOLD) signal, an indirect marker related to local hemodynamic changes accompanying neuronal activity. Researchers examine differences between specified tasks or conditions, such as responses to a cue paired with an aversive outcome versus a safety cue. A stronger BOLD response does not automatically mean that neurons fired more in a simple one-to-one manner, that fear was consciously experienced, or that a region caused the observed behavior.
An especially important limitation is reverse inference: if an area participates in several mental processes, observing activity there cannot uniquely identify one of those processes. Poldrack's methodological analysis explains why the validity of such an inference depends on a region's selectivity, the evidence base, and the statistical model Poldrack, 2011. Amygdala fMRI is also sensitive to small-region segmentation, physiological noise, spatial smoothing, experimental timing, habituation, and design choices. For the underlying methods and their limitations, see Brain Imaging Methods: What MRI, fMRI, PET, and EEG Measure.
EEG, MEG, PET, and psychophysiology
Scalp electroencephalography measures differences in electric potential arising chiefly from coordinated neural currents. Magnetoencephalography measures associated magnetic fields. Both offer high temporal resolution but make localization of activity in a deep and relatively small structure such as the amygdala challenging. They may characterize distributed timing and network responses without directly recording a named amygdala neuron.
Positron emission tomography detects radiation from an administered tracer and can investigate aspects of metabolism or molecular targets depending on tracer selection. It does not provide a general camera view of emotions. Skin conductance, startle electromyography, heart rate, pupil dilation, eye tracking, and behavior measure different peripheral or performance outcomes. These techniques become most informative when researchers specify which mechanism each variable is intended to index.
Lesions, intracranial recordings, and targeted modulation
Naturally occurring lesions can strengthen causal inference about whether a structure contributes to an ability, though affected patients are rare, lesions may be heterogeneous, and long-term neural reorganization complicates interpretation. Intracranial recordings in selected clinical circumstances give higher spatial or temporal access to local electrical signals, but populations are not broadly representative. Animal studies can use cell-specific recording or intervention techniques to test fine circuit hypotheses; the exact results require human validation before clinical translation.
Targeted brain modulation, including the recent ultrasound conditioning work, adds a different kind of intervention evidence. Modulating a region during an experiment is not the same as establishing an approved treatment for a disorder. Device targeting, effects on neighboring circuits, safety, and long-term outcomes remain separate questions.
What happens if the amygdala is damaged?
Consequences depend on whether damage is unilateral or bilateral, the nuclei involved, the age at injury, co-occurring damage, and the task assessed. Documented patterns include changes in recognition and use of threat-related facial information, difficulty acquiring some conditioned defensive reactions, and altered social or emotional learning. Rare cases of extensive bilateral damage have been particularly informative because they help distinguish retained abilities from impaired ones.
A lesion does not necessarily eliminate all fear, abolish every emotion, or erase ordinary memory. Evidence from panic induction after bilateral damage establishes that some highly aversive experiences can still occur. Human adaptation and distributed neural pathways also matter. In everyday clinical practice, deficits require examination in relation to the person's broader neurologic and functional presentation.
Potential causes of amygdala abnormalities or damage include certain neurologic conditions, epilepsy-related pathology or surgery, uncommon genetic or metabolic diseases, and other structural lesions. Their diagnosis and prognosis belong to neurologic assessment, with attention to symptoms, imaging, history, and other findings rather than assumptions derived from a single amygdala measurement.
Is an "overactive amygdala" a medical diagnosis?
The phrase "overactive amygdala" describes a loose popular interpretation of selected research findings, not a standalone clinical diagnosis. Group-average differences in threat-related fMRI responses have been examined in anxiety, posttraumatic stress, depression, and other conditions. Results vary by task, population, treatment, medication exposure, developmental stage, analytic strategy, and the precise outcome measured.
A laboratory association does not establish that a particular person's amygdala is continuously active, chronically injured, or responsible for every symptom. No validated routine individual brain-scan test uses amygdala activation by itself to diagnose an anxiety disorder, PTSD, or another psychiatric condition. Clinical assessment relies on symptoms, duration, functional impact, history, and established diagnostic procedures.
The amygdala is relevant to research on mechanisms implicated in some disorders, but the diagnosis and treatment of those conditions are distinct editorial subjects. For an evidence-focused account of clinical threat-processing findings, see Anxiety and the Brain and How PTSD Affects the Brain.
Can you calm or reset your amygdala?
People can learn skills that change their responses to threat, stress, and emotional situations. Reappraisal, attention shifts, context learning, graded exposure in appropriate clinical care, sleep, exercise, and stress-management practices may influence distributed systems implicated in emotion and learning. Describing these changes as directly "switching off," "detoxing," or permanently "resetting" the amygdala misrepresents what is measured.
For someone struggling with anxiety, panic, or trauma-related symptoms, choosing a care approach should depend on a clinical evaluation and treatment evidence for the actual condition. A sensation of calm after a breathing exercise is a meaningful subjective outcome; it does not prove that a specific nucleus has been deactivated. Likewise, a single fMRI change following an intervention does not establish durable recovery or identify the unique mechanism of benefit.
The practical value of amygdala research is a more precise understanding of emotional learning and context-sensitive behavior. It helps formulate experimental questions, but it does not authorize self-diagnosis from viral brain images or the marketing of unvalidated "amygdala hacks."
What the scientific evidence establishes — and what remains debated
The strongest converging evidence establishes that the amygdala comprises multiple functional circuits involved in learning about significant stimuli, coordinating selected defensive responses, processing reward-related associations, influencing attention, and modulating emotional memory. Anatomical evidence, experimental conditioning, human lesion studies, and comparative work broadly support these conclusions.
Still under investigation are exact computational roles across subdivisions and species; the degree to which cells represent valence, salience, prediction, uncertainty, or combinations of these variables; causal details of large-scale interactions; and the relationship between defensive circuitry and conscious feeling. Contemporary studies with selective cellular recording or brain stimulation help discriminate models, but individual experiments do not resolve the whole field.
Some familiar claims are contradicted by available evidence. The amygdala is not exclusively a fear center; an amygdala response is not a direct readout of someone's fear; removing the amygdala does not universally remove panic; and emotional regulation is not a simple battle between two brain regions. Another common misconception is that a human emotional experience can be directly inferred from a rodent freezing response. These are different measurements that require careful translation.
Frequently asked questions
What is the amygdala's main function?
Its most defensible broad description is contributing to learning and using the significance of events and cues, including threats, rewards, uncertainty, and social information. Threat conditioning is especially well studied; coordinated responses arise through interaction with other circuits.
Is the amygdala responsible for fight or flight?
It participates in pathways that influence defensive behavior and autonomic regulation, but fight, flight, freezing, vigilance, and other reactions involve hypothalamic, brainstem, cortical, and peripheral systems. Different species and contexts also show different response patterns.
Does the amygdala control anger?
Amygdala circuits can contribute to processing relevant stimuli and organizing motivated behavior, including certain responses associated with anger or aggression. Anger is a complex affective and social phenomenon supported by interacting systems. A scan of the amygdala cannot identify anger by itself.
Does the amygdala control happiness or pleasure?
It contributes to some reward-related learning and valuation, especially associations among stimuli and outcomes. Pleasure and happiness are broader phenomena with multiple neural and psychological contributors. Describing the amygdala as a pleasure or fear switch is inaccurate.
What is the difference between the amygdala and hippocampus?
The amygdala is particularly involved in assigning and learning significance and in influencing response systems and emotional-memory strength. The hippocampus is centrally involved in contextual, relational, episodic, and spatial representations. They interact extensively rather than dividing every experience into separate emotion and memory compartments.
Can you live without an amygdala?
Rare people with extensive bilateral damage have survived and retain many mental and emotional abilities, while showing selective alterations in fear learning, social perception, and other functions. Outcomes vary, and individual case findings cannot be generalized to every possible injury.
Is a larger amygdala better or worse?
Neither conclusion follows from size alone. Structural differences have many determinants and do not provide a reliable direct score of emotional strength, resilience, anxiety, or personality. Interpretation requires appropriate population comparisons and clinical context.
Can amygdala activity diagnose PTSD or anxiety?
An amygdala BOLD response is a research measure that may be informative at the group level in a defined paradigm. It is not an established standalone diagnostic biomarker for an individual. Appropriate evaluation uses clinical criteria and evidence-based assessment.
Can the amygdala change through experience?
Yes. Synaptic and circuit plasticity are central to learning in the amygdala and its connected networks. The extent and consequences of change depend on the type of experience, the population studied, and the measures obtained. Brain plasticity does not imply that any particular self-help method can permanently reprogram a specific nucleus.
Do animals and humans have the same amygdala?
Both share major evolutionary and anatomical organizational features, enabling comparative research. Homologous structures, cell populations, connectivity, behavioral repertoires, and conscious reports are not identical. Species-specific evidence should be identified as such.
Related Articles
For condition-specific neurobiology, read Anxiety and the Brain: Threat Processing, Learning, and What Neuroimaging Can Show, How PTSD Affects the Brain: Neurobiology, Evidence, and Misconceptions, Depression and the Brain: What Neuroscience Actually Shows, and How Childhood Trauma Affects the Brain. These articles retain their existing clinical canonical ownerships.
References
Adolphs, R., Gosselin, F., Buchanan, T. W., Tranel, D., Schyns, P., & Damasio, A. R. (2005). A mechanism for impaired fear recognition after amygdala damage. Nature, 433, 68–72. https://doi.org/10.1038/nature03086
Aksoy-Aksel, A., Ferraguti, F., Holmes, A., Lüthi, A., & Ehrlich, I. (2025). Amygdala intercalated cells form an evolutionarily conserved system orchestrating brain networks. Nature Neuroscience, 28, 234–247. https://doi.org/10.1038/s41593-024-01836-8
Baxter, M. G., & Murray, E. A. (2002). The amygdala and reward. Nature Reviews Neuroscience, 3, 563–573. https://doi.org/10.1038/nrn875
Feinstein, J. S., Adolphs, R., Damasio, A., & Tranel, D. (2011). The human amygdala and the induction and experience of fear. Current Biology, 21, 34–38. https://doi.org/10.1016/j.cub.2010.11.042
Feinstein, J. S., Buzza, C., Hurlemann, R., et al. (2013). Fear and panic in humans with bilateral amygdala damage. Nature Neuroscience, 16, 270–272. https://doi.org/10.1038/nn.3323
Klumpers, F., Morgan, B., Terburg, D., Stein, D. J., & van Honk, J. (2015). Impaired acquisition of classically conditioned fear-potentiated startle reflexes in humans with focal bilateral basolateral amygdala damage. Social Cognitive and Affective Neuroscience, 10, 1161–1168. https://doi.org/10.1093/scan/nsu164
Kong, M.-S., Ancell, E., Witten, D. M., & Zweifel, L. S. (2025). Valence and salience encoding in the central amygdala. eLife, 13, RP101980. https://doi.org/10.7554/eLife.101980.3
LeDoux, J. E., & Pine, D. S. (2016). Using neuroscience to help understand fear and anxiety: A two-system framework. American Journal of Psychiatry, 173, 1083–1093. https://doi.org/10.1176/appi.ajp.2016.16030353
McGaugh, J. L. (2004). The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annual Review of Neuroscience, 27, 1–28. https://doi.org/10.1146/annurev.neuro.27.070203.144157
Meijer, S., Carpino, E., Kop, B. R., et al. (2026). The human amygdala in threat learning and extinction. Science Advances, 12, eaea8233. https://doi.org/10.1126/sciadv.aea8233
Pessoa, L. (2010). Emotion and cognition and the amygdala: From "what is it?" to "what's to be done?" Neuropsychologia, 48, 3416–3429. https://doi.org/10.1016/j.neuropsychologia.2010.06.038
Pessoa, L., & Adolphs, R. (2010). Emotion processing and the amygdala: From a "low road" to "many roads" of evaluating biological significance. Nature Reviews Neuroscience, 11, 773–782. https://doi.org/10.1038/nrn2920
Poldrack, R. A. (2011). Inferring mental states from neuroimaging data: From reverse inference to large-scale decoding. Neuron, 72, 692–697. https://doi.org/10.1016/j.neuron.2011.11.001
Sander, D., Grafman, J., & Zalla, T. (2003). The human amygdala: An evolved system for relevance detection. Reviews in the Neurosciences, 14, 303–316. https://doi.org/10.1515/revneuro.2003.14.4.303
Sias, A. C., Jafar, Y., Goodpaster, C. M., et al. (2024). Dopamine projections to the basolateral amygdala drive the encoding of identity-specific reward memories. Nature Neuroscience, 27, 728–736. https://doi.org/10.1038/s41593-024-01586-7
Wang, Y., Olsson, S., Lipp, O. V., & Ney, L. J. (2024). Renewal in human fear conditioning: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews, 159, 105606. https://doi.org/10.1016/j.neubiorev.2024.105606