Anxiety Sensitivity: The Fear of Anxiety Itself

Most people experience anxiety as unpleasant. But for some, the unpleasantness goes further: they are not just anxious but frightened of what the anxiety sensations themselves might mean. A racing heart is not merely uncomfortable — it signals a heart attack. A moment of confusion is not fatigue — it is the beginning of losing one’s mind. This fear of anxiety itself is what researchers call anxiety sensitivity, and it has been studied as a distinct construct for four decades.

What it is and what it is not

Anxiety sensitivity was formally described by Reiss and colleagues in 1986 as the belief that anxiety sensations have harmful consequences beyond their immediate discomfort [1]. It is conceptually distinct from trait anxiety, which is a general tendency to feel anxious across situations. The distinction matters: two people can experience the same level of anxiety, but the one with high anxiety sensitivity interprets those symptoms as dangerous and escalates.

The construct has three dimensions, reflected in the subscales of the Anxiety Sensitivity Index-3 (ASI-3) [2]:

• Physical concerns: “When my stomach is upset, I worry that I might be seriously ill.” The fear is that bodily sensations signal medical catastrophe.
• Cognitive concerns: “When I cannot keep my mind on a task, I worry that I might be going crazy.” The fear is that mental symptoms signal loss of control or insanity.
• Social concerns: “It is important to me not to appear nervous.” The fear is that visible anxiety will lead to social rejection or humiliation.

These dimensions have different associations with different disorders, which is part of what makes the construct useful.

The evidence for specificity

A meta-analysis of 172 studies examined how anxiety sensitivity relates to the various anxiety and depressive disorders [3]. The findings supported partial specificity:

• Physical concerns were most strongly associated with panic disorder and agoraphobia.
• Cognitive concerns were associated with depression and generalized anxiety.
• Social concerns were associated with social anxiety disorder.

Overall, anxiety sensitivity was a stronger predictor of panic disorder than of other anxiety conditions, even after controlling for general trait anxiety. This partial specificity is what justifies treating anxiety sensitivity as a separate construct rather than just another name for being anxious.

Prospective evidence

The most important question for a risk factor is whether it predicts future problems, not just correlates with current ones. A prospective study followed 404 young adults over two years and found that baseline anxiety sensitivity predicted the occurrence of future panic attacks, even after controlling for prior panic history and trait anxiety [4]. Those with the highest anxiety sensitivity scores were several times more likely to experience panic attacks over the follow-up period.

This is not proof of causation — other unmeasured factors could drive both anxiety sensitivity and panic. But it is more than cross-sectional correlation, and it is consistent with the theoretical model: if you fear the sensations of anxiety, encountering those sensations becomes a trigger for escalation.

Can it be changed?

If anxiety sensitivity is a modifiable risk factor rather than a fixed trait, reducing it might prevent panic. Several lines of evidence suggest it is modifiable:

Interoceptive exposure. Exercises that deliberately induce anxiety sensations — spinning in a chair (dizziness), breathing through a straw (breathlessness), running in place (racing heart) — have been shown to reduce anxiety sensitivity. A brief single-session intervention combining psychoeducation with interoceptive exposure reduced anxiety sensitivity scores and, at six-month follow-up, reduced the incidence of anxiety pathology compared with a health information control [6].

Exercise. A small randomized trial found that a two-week exercise program (six sessions of 20 minutes of aerobic exercise) reduced anxiety sensitivity significantly more than a waitlist control [5]. The proposed mechanism is repeated, non-threatening exposure to the physical sensations of exertion — sweating, elevated heart rate, breathlessness — which teaches the person that these sensations are not dangerous. See exercise for anxiety evidence for the broader literature on exercise and anxiety.

CBT for panic. Standard CBT for panic disorder includes interoceptive exposure as a core component, and anxiety sensitivity typically decreases during successful treatment. Whether the reduction in anxiety sensitivity is a mechanism of change or a consequence of improvement is debated.

The measurement issue

The ASI and its revisions are self-report measures. They ask people to rate how much they fear various anxiety symptoms. Self-report measures are vulnerable to the same issues discussed in self-report versus clinician-rated measures: social desirability, recall bias and the assumption that people can accurately report their own beliefs.

There is also a conceptual circularity risk. If anxiety sensitivity is measured by asking whether anxiety symptoms are frightening, and the outcome is anxiety or panic (which involves finding anxiety symptoms frightening), the predictor and the outcome are uncomfortably close to measuring the same thing. Researchers have addressed this partly through prospective designs and by showing that anxiety sensitivity predicts panic specifically rather than distress generally, but the concern has not fully disappeared.

What remains open

• Whether anxiety sensitivity is genuinely causal or simply an early marker of a broader vulnerability is unresolved. Experimental manipulations of anxiety sensitivity (like the brief interventions above) that subsequently reduce panic would support a causal role, and some evidence points in this direction, but large-scale prevention trials are lacking.
• How the three dimensions develop — whether through direct experience (a parent who modeled fear of bodily sensations), learning, temperament or some combination — is not well mapped.
• Whether anxiety sensitivity is best understood as a continuum (everyone has some level) or a taxon (a qualitatively distinct group) has been tested with mixed results.

When to look for more support. If you recognize the pattern described here — fearing what your anxiety symptoms mean, avoiding activities that produce bodily sensations, interpreting physical symptoms as dangerous — talk to a GP or psychologist. Panic disorder responds well to CBT, particularly with interoceptive exposure. If you are in crisis or think you may be having a medical emergency, seek emergency care. For mental health crises: call or text 988 (US) or Samaritans on 116 123 (UK and Ireland).