What Causes Anxiety: The Neuroscience and Psychology of Fear

What Is Anxiety?

Anxiety is a future-oriented emotional state characterized by apprehension about potential threats, physiological arousal, and behavioral readiness for protective action - distinct from fear, which is a response to an immediate, present-moment danger.

It is produced by biological systems, centered on the amygdala and HPA axis, that evolved to detect and respond to uncertain threat, and it becomes clinically significant when it is disproportionate, persistent, or functionally impairing.

Approximately 30% of adults will meet criteria for an anxiety disorder at some point in their lives, making it the most prevalent category of mental health condition globally.

In 1871, Charles Darwin noticed something strange about himself. During a visit to the London Zoo, he resolved to stand perfectly still in front of a puff adder's glass cage and allow the snake to strike. He knew the glass was between them. He knew there was no danger. The strike came, and Darwin jumped backward.

"My will and reason were powerless against the imagination of a danger which had never been experienced," he wrote in The Expression of the Emotions in Man and Animals.^[8]

He was describing, with characteristic precision, what would take another century of neuroscience to explain: that fear responses are generated by systems that operate faster than conscious deliberation, are not subject to rational override, and were built for an environment very different from the one we now inhabit.

Anxiety is one of the most common human experiences - approximately 30% of adults will meet criteria for an anxiety disorder at some point in their lives, making it the most prevalent category of mental health condition globally.

It is also one of the most misunderstood: often dismissed as excessive worry, weakness, or irrationality, when in fact it reflects the operation of biological systems doing precisely what evolution designed them to do, in a modern environment those systems were not designed for.

"Anxiety is the dizziness of freedom." - Soren Kierkegaard

Key Definitions

Fear - An immediate emotional response to a specific, concrete, present threat. Time-locked to the moment; resolves when the threat passes. Mediated primarily by the amygdala and its downstream connections.

Anxiety - A future-oriented emotional state characterized by apprehension about potential threats, uncertainty, or negative outcomes. Not tied to a specific present-moment threat; anticipatory and often diffuse.

Amygdala - An almond-shaped subcortical structure in the medial temporal lobe, the brain's primary threat-detection and alarm system. Rapidly processes sensory information for threat relevance, often before conscious awareness; activates the HPA axis and sympathetic nervous system.

Bed nucleus of the stria terminalis (BNST) - A structure closely connected to the amygdala that mediates sustained, diffuse anxiety states (as opposed to acute fear responses). Increasingly understood as the substrate for anxious apprehension rather than acute fear.

HPA axis - The hypothalamic-pituitary-adrenal axis: the stress-response system that produces cortisol. Activated by amygdala threat detection; mobilizes energy, heightens alertness, and prepares the body for threat response.

Extinction learning - The process by which a conditioned fear response diminishes through repeated exposure to the feared stimulus without the aversive outcome. Does not erase the original fear memory but creates a competing safety memory. The neural basis of exposure therapy.

Interoception - The perception of one's own internal bodily states - heartbeat, breathing, muscle tension. Abnormal interoception (hypervigilance to internal sensations) is central to panic disorder.

Safety behaviors - Actions taken to reduce anxiety in feared situations (carrying medication "just in case," checking repeatedly, avoiding eye contact in social situations). Provide short-term relief but maintain anxiety long-term by preventing extinction learning.

Behavioral inhibition - A temperament characterized by wariness, withdrawal, and physiological reactivity to novelty, particularly prominent in early childhood. A significant predictor of later anxiety disorders.

The Two Pathways: How the Brain Processes Threat

Joseph LeDoux's research at NYU in the 1980s and 1990s produced one of the most influential models in affective neuroscience: the two-pathway model of fear processing.^[2]

When a potentially threatening sensory stimulus enters the brain, it is processed simultaneously along two routes:

The low road: A rapid, subcortical pathway from the thalamus directly to the amygdala, bypassing the cortex entirely. This pathway is fast (~12ms) and coarse-grained - it can trigger a fear response based on rough, incomplete information.

LeDoux famously demonstrated this with rats: severing the direct thalamo-amygdala pathway eliminated the conditioned fear response, even when cortical processing remained intact.^[1]

The high road: A slower pathway from the thalamus to the sensory cortices, then to the prefrontal cortex, then to the amygdala. This route takes longer (~30-40ms) but processes richer, more accurate information, allowing the cortex to evaluate whether the threat is real, in context, and proportionate.

The low road produces the "jump" before you register what made you jump. Darwin's automatic recoil from the puff adder had already completed through the low road before his cortex finished constructing the conscious representation of "glass barrier, no actual danger."

In anxiety disorders, the low road is hyperactive and the high-road regulatory override is impaired - the alarm fires easily and the cortical braking mechanism is inefficient.

The Amygdala: More Than a Fear Center

The amygdala has been called "the brain's fear center" in countless popular accounts, but this is an oversimplification. The amygdala is better understood as a general salience detector and emotional learning hub - it responds strongly to anything relevant to survival (threat, reward, novel stimuli), not to threat alone.

What the amygdala does in anxiety:

Threat Detection and Alarm

The amygdala's basolateral complex receives sensory input and rapidly evaluates its threat relevance - activating when it detects anything that resembles a previously feared stimulus, an ambiguous social signal, or anything learned to predict danger.

Neuroimaging studies consistently find heightened amygdala activation in anxiety disorder patients compared to controls:^[3]

• Greater response to threatening faces (fear, anger)
• Greater response to ambiguous stimuli (neutral faces are often processed as threatening)
• Greater response at lower threat intensities (the alarm threshold is lower)
• Slower return to baseline after threat stimuli resolve

This heightened sensitivity is not simply a psychological trait - it is a neurobiological difference measurable in the scanner.

Downstream Consequences

When the amygdala activates:

• The hypothalamus triggers the HPA axis: corticotropin-releasing hormone (CRH) → ACTH ? cortisol release from the adrenal cortex
• The sympathetic nervous system activates: elevated heart rate, blood pressure, respiratory rate; dilation of pupils; redirection of blood flow to large muscle groups; inhibition of digestion
• The locus coeruleus releases norepinephrine throughout the brain, increasing vigilance and arousal
• Attention is directed toward potential threats (threat-attention bias - anxious individuals automatically attend to threatening stimuli in their environment)

This full-body mobilization is the fight-or-flight response - highly adaptive in response to a genuine physical threat, expensive and disruptive when maintained chronically in response to abstract future threats.

The Prefrontal Failure: Why Reason Can't Always Override Anxiety

If anxiety is driven by amygdala hyperactivity, why can't people simply reason their way out of it?

The prefrontal cortex (PFC) - particularly the ventromedial PFC (vmPFC) - is the primary brake on amygdala activity. Through direct inhibitory projections, the vmPFC can reduce amygdala activity when cognitive evaluation determines the threat is not real or not proportionate.

In non-anxious individuals, this PFC regulatory circuit functions efficiently: the cortex evaluates "this is not dangerous," signals the amygdala to stand down, and the fear response diminishes.

In anxiety disorders, PFC-amygdala connectivity is functionally reduced. Neuroimaging shows weaker functional connectivity between vmPFC and amygdala in anxious individuals, meaning the cognitive "this isn't actually dangerous" signal is less effective at damping the amygdala's alarm response.

This is why telling an anxious person "there's nothing to worry about" rarely works. It is not that the anxious person hasn't considered this. It is that the regulatory circuit that would translate this cognitive evaluation into reduced amygdala activity is not functioning normally.

Darwin knew cognitively that the glass was between him and the snake. His cortex had perfectly accurate information. His low road didn't care.

Generalized Anxiety: When the Alarm Has No Target

Generalized Anxiety Disorder (GAD) - characterized by chronic, difficult-to-control worry across multiple domains - represents anxiety divorced from any specific, identifiable threat. The patient worries about work, health, money, relationships, safety - an ever-rotating roster of potential catastrophes.

The worry in GAD has a distinctive structure. It is:

• Future-oriented: Not responding to present threats but to imagined future ones
• Verbal-linguistic: Primarily cognitive rumination rather than vivid imagery (which is important - verbal worry may actually suppress the visceral distress of anxiety, making it self-reinforcing)
• Intolerance of uncertainty-driven: GAD patients show specific elevation in intolerance of uncertainty (IU) - the belief that uncertainty itself is unacceptable, threatening, and indicative of imminent negative outcomes

Michel Dugas and colleagues at the University of Quebec have developed the intolerance of uncertainty model of GAD: the core psychological problem is not excessive fear of specific outcomes but an excessive need for certainty that the future will be manageable.^[6]

Worry is an attempt - ultimately futile - to achieve certainty through mental simulation of potential problems.

The vmPFC-hippocampus-amygdala circuitry is specifically involved in uncertainty processing: highly uncertain outcomes produce greater amygdala activation than certain negative outcomes in many studies. Unpredictability, not just negativity, drives the brain's alarm system.

Social Anxiety: The Most Common Anxiety Disorder

Social anxiety disorder (SAD) - intense fear of social situations involving potential scrutiny, embarrassment, or negative evaluation - affects approximately 12% of people over their lifetime, making it the most common anxiety disorder by lifetime prevalence.^[4]

The feared outcome in social anxiety is social rejection or humiliation - evolutionarily ancient fears. For social primates, exclusion from the group was a genuine survival threat; social evaluation anxiety is not irrational but may be miscalibrated to modern social environments.

The Self-Focus Trap

Clark and Wells' cognitive model of social anxiety proposes that a key maintaining mechanism is self-focused attention: in anxiety-provoking social situations, the socially anxious person turns attention inward, monitoring their own perceived performance, anxiety symptoms, and social behavior from a critical, self-observing perspective.^[5]

This self-monitoring is self-defeating: it consumes cognitive resources that should be allocated to actual social engagement (listening, responding, processing the conversation), and creates a distorted "internal observer" perspective that exaggerates visible anxiety symptoms.

The person believes they appear much more anxious than they actually do.

Neuroimaging of social anxiety shows heightened activity in the medial PFC (associated with self-referential processing) and amygdala during social evaluation tasks, with reduced activity in regions associated with actual social processing (the "social brain" network).

Panic Disorder: Fear of Fear Itself

Panic disorder presents a distinctive phenomenology: the feared object is the internal physical sensations of anxiety itself.

A panic attack - sudden intense fear with physical symptoms including racing heart, shortness of breath, dizziness, chest pain, and derealization - is intensely alarming partly because its physical symptoms mimic those of cardiac arrest.

The maintaining mechanism in panic disorder is the interoceptive fear cycle:

1. Harmless body sensation (slightly elevated heart rate from caffeine, mild dizziness)
2. Catastrophic interpretation ("I'm having a heart attack / I'm going to faint / I'm dying")
3. Anxiety response to the interpretation
4. Anxiety produces more intense physical symptoms (heart rate accelerates further, hyperventilation begins)
5. Amplified symptoms confirm the catastrophic interpretation ? panic attack

The core cognitive pathology in panic is anxiety sensitivity - the belief that anxiety symptoms are themselves harmful or dangerous. This predicts panic disorder more strongly than anxiety frequency.

Interoception research, led by Hugo Critchley and others, finds that panic disorder patients show hypersensitivity to interoceptive signals (heartbeat, respiratory changes) - the internal alarm has a lower threshold. The anterior insula - the primary interoceptive cortex - shows altered activity and connectivity in panic disorder.

Why Some People Develop Anxiety Disorders

The Diathesis-Stress Model

No single cause produces anxiety disorders. The most supported framework is the diathesis-stress model: a person carries a predisposing vulnerability (diathesis) that, under sufficient environmental stress, tips into a clinical disorder.

Genetic vulnerability: Twin studies find heritability of 30-50% for anxiety disorders. The strongest genetic factor is not anxiety-specific but is the neuroticism dimension of personality - a general trait predisposing to negative affect, worry, and emotional reactivity.

GWAS studies have identified many common variants of small effect in genes related to HPA axis function, serotonin signaling, and GABA-A receptor function.

Early life experience: Adversity in childhood - abuse, neglect, instability, loss, parental anxiety - calibrates the stress-response system. Chronically stressed children often develop a hyperreactive HPA axis that remains elevated in its threat-response sensitivity into adulthood.

The amygdala-PFC regulatory circuitry develops partly through experience; an environment with frequent, unpredictable threats may produce a circuit that learns to be chronically vigilant.

Behavioral inhibition: A temperament characterized in infancy and early childhood by wariness, clinging, withdrawal from novel people and situations, and physiological reactivity (elevated heart rate, cortisol) to mild novelty.

Approximately 15% of children show this temperament; it is moderately heritable and predicts social anxiety and other anxiety disorders in adolescence and adulthood.

The conditioning history: Many anxiety responses are learned - through direct classical conditioning (experiencing something frightening in a particular context), observational conditioning (watching someone else be frightened), or informational transmission (being told repeatedly that something is dangerous).

A child who grows up with an anxious parent who models the world as threatening may acquire anxiety responses to situations that are not objectively threatening.

What Anxiety Does to the Body Over Time

Acute anxiety is uncomfortable; chronic anxiety is damaging.

Cortisol - the primary stress hormone - is acutely adaptive and chronically toxic. Hippocampal neurons, which express high densities of cortisol receptors, are specifically damaged by chronically elevated cortisol, explaining the documented hippocampal volume reductions in people with chronic anxiety and PTSD.

The relationship between anxiety and cardiovascular disease is epidemiologically documented: anxiety disorders are associated with approximately 25-40% increased risk of cardiovascular events, beyond what is explained by lifestyle factors. The mechanism likely involves chronic sympathetic activation, inflammation, and HPA dysregulation.

The Evolutionary Mismatch

Anxiety mechanisms were calibrated for an environment of concrete, physical, short-duration threats - the puff adder behind the glass, the rival in the tribal hierarchy, the storm on the horizon. The fight-or-flight response resolves when the threat passes; the body returns to baseline; the stress hormones clear.

Modern life provides a different threat profile: chronic, abstract, social, and often unresolvable.

Financial insecurity, career uncertainty, social media comparison, existential threats (climate, political instability) - these are real stressors, but they don't resolve through physical action, and the body's stress response keeps activating without the discharge that physical response would provide.

The result is what stress researchers call allostatic load - the cumulative wear on biological systems from chronic activation of the stress response in the absence of resolution.

The same systems that protected our ancestors from predators are, in modern environments, maintaining us in a low-grade state of physiological mobilization that compounds over years into measurable health consequences.

What Actually Helps

Exposure therapy remains the most effective intervention for anxiety disorders, with large effect sizes (d=0.9-1.5) across phobias, social anxiety, panic disorder, and OCD.^[7]

The mechanism: repeated confrontation with feared stimuli or situations, without safety behaviors, allows extinction learning - new "safe" memories are formed that compete with the fear association.

The goal is not to eliminate the fear memory (which is impossible) but to create a competing safety memory that wins out in most contexts.

Cognitive-behavioral therapy (CBT) addresses the cognitive distortions (catastrophic interpretation, threat overestimation, intolerance of uncertainty) that maintain anxiety. Combined with exposure, CBT has long-term outcomes superior to medication alone.

Aerobic exercise: Meta-analyses find effects comparable to SSRIs for anxiety reduction. Mechanisms include BDNF production (which supports hippocampal neurogenesis and may reverse stress-related damage), endocannabinoid release, and direct HPA axis modulation.

SSRIs/SNRIs: Effective for most anxiety disorders, working by gradually downregulating amygdala threat sensitivity over 4-8 weeks. Not habit-forming; best used in combination with psychotherapy for durable results.

Acceptance-based approaches (ACT): Rather than attempting to control or eliminate anxiety, acceptance and commitment therapy builds tolerance for anxiety while committing to valued behavior - changing the relationship to anxiety rather than the anxiety itself.

Particularly useful for people whose attempts to control anxiety have themselves become a maintaining factor.

For related concepts, see how to manage anxiety, what causes depression, how stress damages the body, and how the teenage brain works.

Sources & Further Reading

1. LeDoux, J. E. (2015). Anxious: Using the Brain to Understand and Treat Fear and Anxiety. Viking.
2. LeDoux, J. E. (1996). The Emotional Brain. Simon & Schuster.
3. Etkin, A., & Wager, T. D. (2007). Functional Neuroimaging of Anxiety. American Journal of Psychiatry, 164(10), 1476-1488. DOI: 10.1176/appi.ajp.2007.07030504
4. Kessler, R. C., et al. (2005). Prevalence, Severity, and Comorbidity of 12-Month DSM-IV Disorders. Archives of General Psychiatry, 62(6), 617-627. DOI: 10.1001/archpsyc.62.6.617
5. Clark, D. M., & Wells, A. (1995). A Cognitive Model of Social Phobia. In R. G. Heimberg et al. (Eds.), Social Phobia: Diagnosis, Assessment, and Treatment (pp. 69-93). Guilford Press.
6. Dugas, M. J., Gagnon, F., Ladouceur, R., & Freeston, M. H. (1998). Generalized Anxiety Disorder: A Preliminary Test of a Conceptual Model. Behaviour Research and Therapy, 36(2), 215-226. DOI: 10.1016/S0005-7967(97)00070-3
7. Hofmann, S. G., & Smits, J. A. J. (2008). Cognitive-Behavioral Therapy for Adult Anxiety Disorders: A Meta-Analysis of Randomized Placebo-Controlled Trials. Journal of Clinical Psychiatry, 69(4), 621-632. DOI: 10.4088/jcp.v69n0415
8. Darwin, C. (1872). The Expression of the Emotions in Man and Animals. John Murray.