The Nocebo Effect in Anxiety Treatment: When Expecting Side Effects Makes Them More Likely

In anxiety medication trials, something counterintuitive happens: participants who receive a placebo pill frequently report side effects. Headaches, nausea, dizziness, insomnia — symptoms they attribute to a medication that contains no active ingredient. This is the nocebo effect, and it is not an oddity of trial design. It is a window into how expectation shapes the experience of treatment.

What the nocebo effect is

The nocebo effect is the negative counterpart of the placebo effect. While placebo effects involve positive expectations producing positive outcomes, nocebo effects involve negative expectations producing negative ones. A patient who reads the list of possible side effects and expects nausea is more likely to experience nausea, regardless of what they swallowed [1].

The mechanisms are psychological and neurobiological. Expectation activates anxiety and attentional vigilance toward the expected symptom, which both lowers the threshold for detecting the sensation and amplifies its perceived intensity. An fMRI study showed that negative expectations about a drug literally modulated its analgesic efficacy in the brain, turning a demonstrably effective opioid into an ineffective one [5].

How large it is in anxiety-relevant trials

Antidepressants are the most commonly prescribed medications for anxiety disorders, and their trials are the most studied for nocebo effects. A meta-analysis of nocebo effects in antidepressant trials for depression found that the average dropout rate due to adverse events in the placebo arm was approximately 5% [2]. Individual symptoms were reported at higher rates: headache in 16–20% of placebo-group participants, nausea in 10–15%, dizziness in 5–10%.

These rates are not trivial. They are high enough to contribute to treatment discontinuation and to complicate the interpretation of side-effect profiles. If 15% of people on placebo report nausea and 25% on the drug report nausea, the drug’s “true” side-effect rate for nausea may be closer to 10% (the excess over placebo) rather than the 25% listed in the package insert.

Strikingly, placebo side-effect profiles tend to mirror the drug they are compared with. In a systematic review of anti-migraine trials, the adverse events reported in the placebo arm matched those expected from the specific drug class being tested [6]. Participants in a triptan trial reported triptan-like side effects on placebo; participants in an anti-inflammatory trial reported anti-inflammatory-like side effects. This suggests that the informed consent process itself generates the nocebo response.

Why anxiety may amplify nocebo

Several features of anxiety could increase susceptibility to nocebo effects:

Heightened body monitoring. People with anxiety, particularly health anxiety and panic disorder, attend more to bodily sensations. Ambiguous sensations (a mild stomach gurgle, slight dizziness on standing) that a non-anxious person would ignore may be noticed, attributed to the medication, and reported as a side effect.

Catastrophic interpretation. The same cognitive tendency that drives anxiety — interpreting ambiguous signals as threatening — applies to medication. A mild headache becomes “the medication is harming my brain.”

Information sensitivity. A meta-analysis of factors contributing to nocebo effects found that the amount and framing of negative information provided about a treatment predicted the magnitude of nocebo responses [3]. Anxious people are more sensitive to threat information generally, which may amplify the effect of reading a side-effect leaflet.

Attentional bias. The attentional biases documented in anxiety disorders (see attention bias modification) could selectively focus attention on bodily sensations consistent with expected side effects.

A systematic review of nocebo predictors found that anxiety, negative expectations and prior adverse treatment experiences were the most consistent predictors of nocebo responding [4].

Clinical implications

The nocebo effect creates a practical dilemma for clinicians. Informed consent requires telling patients about possible side effects. But providing that information increases the likelihood of experiencing them. How should this be navigated?

Several approaches have been proposed:

• Positive framing. Leading with efficacy information (“Most patients tolerate this medication well and notice improvement within a few weeks”) before listing side effects, rather than opening with a catalog of possible harms.
• Contextualized information. Providing rates (“About 1 in 10 people experience nausea”) rather than lists, so patients can calibrate the actual probability.
• Expectation checking. Asking what the patient expects and addressing specific fears. A patient who had a bad experience with a previous SSRI may carry expectations that do not apply to the current medication.
• Minimizing unnecessary warnings. Distinguishing between common side effects that patients need to know about and rare serious effects that require medical attention, rather than listing every reported adverse event.

None of these approaches has been tested in randomized trials specifically in anxiety populations.

Implications for reading trial results

The nocebo effect means that side-effect rates in drug trials cannot be read at face value. The informative comparison is always drug versus placebo:

1. What percentage of the drug group reported a given side effect?
2. What percentage of the placebo group reported the same side effect?
3. The difference is the best estimate of the drug’s true side-effect rate.

Trial reports that list only absolute side-effect rates without the placebo comparison are misleading. For the related issue of how large placebo responses are in anxiety trials and what drives them, see the placebo response in anxiety trials.

When to look for more support. If you are experiencing side effects that concern you, talk to the prescribing doctor before stopping medication. Some early side effects resolve in the first weeks. If you have stopped or want to stop an anxiety medication, do so with medical guidance — abrupt discontinuation of some medications can itself cause symptoms. If you are in crisis, call or text 988 (US) or Samaritans on 116 123 (UK and Ireland).