Psychology · 6 min read
How Expectations and Learning Can Alter Pain
How expectation, conditioning, descending modulation, and endogenous opioids contribute to bounded placebo analgesia.

How Expectations and Learning Can Alter Pain
Pain is not a direct meter reading of tissue damage. It is an experience produced when the nervous system integrates sensory input with attention, threat, context, memory, and expectations. This does not make pain imaginary. It means that an inert treatment presented as helpful can sometimes reduce reported pain through learned and expected effects, while negative expectations can increase it. These changes are called placebo analgesia and nocebo hyperalgesia.
The evidence is strongest in controlled experimental pain and in some short-term clinical outcomes. Mechanisms vary. Endogenous opioid signaling contributes in certain paradigms; conditioning and non-opioid systems contribute in others. Brain imaging identifies networks associated with expectation and pain evaluation, but it cannot by itself show that a scan reflects less nociceptive input. Placebo effects do not reliably remove disease causes or substitute for effective treatment.
Pain includes input, modulation, and evaluation
Nociceptors detect potentially damaging mechanical, thermal, or chemical events. Signals travel through the spinal cord toward brainstem, thalamic, sensory, affective, and cognitive systems. At several levels, neural activity can be amplified or inhibited. Descending pathways from the brain can influence spinal processing. Attention and expectations change how incoming signals are interpreted and regulated.
Pain and nociception are therefore related but not identical. People can have tissue damage with little pain, persistent pain after healing, or pain without a clearly identifiable lesion. The International Association for the Study of Pain defines pain as a sensory and emotional experience and emphasizes that a person's report should be respected. Contextual modulation is part of biology, not evidence of deception.
A placebo in a trial is often an inert pill or sham procedure designed to resemble treatment. The “placebo response” is the total change observed after it, which includes natural fluctuation, regression toward the mean, reporting effects, and the treatment context. The “placebo effect” more narrowly refers to change caused by that context compared with an appropriate no-treatment or different-context control. Confusing response with effect exaggerates the causal contribution.
Expectations and conditioning can work together
In an expectation experiment, researchers may apply identical heat stimuli to two skin sites but tell participants that one site received an analgesic cream. Lower pain reports at the “treated” site estimate an effect of the instruction and ritual. Stronger designs conceal stimulus changes, randomize conditions, and separate suggestion from actual intensity.
Conditioning adds learning. During training, a cue is paired with genuinely lower pain. Later, identical stimuli are delivered under the learned cue and a control cue. Participants may continue to report less pain under the cue associated with relief. Fabrizio Benedetti and colleagues have shown that verbal expectation and conditioning can produce different pharmacological patterns, suggesting that placebo analgesia is not one mechanism.
Clinical encounters combine these elements. The appearance of a pill, prior relief, clinician communication, price cues, and the act of treatment can shape predictions. Open-hidden studies demonstrate the contribution of awareness: the same analgesic can have a larger measured effect when a patient knows it is being administered than when an infusion pump delivers it without an explicit signal. That difference does not mean the drug is inert. It indicates that pharmacology and context can add or interact.
Expectations are not voluntary commands. Telling a person to “think positively” does not guarantee analgesia, and failed relief is not the patient's fault. Effects vary between people and episodes. Attempts to identify a stable “placebo responder” have had limited success because response depends on treatment, learning history, outcome, and measurement noise.
Endogenous opioids explain some paradigms
In 1978, Jon Levine, Newton Gordon, and Howard Fields reported that naloxone, an opioid antagonist, reduced placebo analgesia after dental surgery. Later studies supported endogenous opioid involvement in some expectation-based effects. Positron-emission tomography research has observed changes in mu-opioid receptor activity during placebo conditions in regions involved in pain and affect.
Naloxone does not block each placebo analgesic effect. Conditioning with non-opioid drugs can recruit other pathways, and different pain models yield different results. A systematic review of early naloxone studies found evidence consistent with opioid mediation but noted small samples and methodological variation. The precise claim is that endogenous opioids can mediate placebo analgesia under particular conditions, not that a single “placebo chemical” explains the phenomenon.
Other candidates include dopamine in reward and learning, cannabinoids in some conditioned responses, and descending brainstem control. These systems interact. Neuropharmacological reviews emphasize that mechanisms in chronic pain are less established than mechanisms in short laboratory experiments.
Imaging locates associations, not a pain switch
Functional MRI studies compare blood-oxygen-level-dependent signals during placebo and control conditions. Early work reported changes in prefrontal cortex, anterior cingulate cortex, insula, thalamus, and periaqueductal gray. These regions participate in expectation, valuation, attention, affect, and descending modulation.
An individual-participant meta-analysis led by Tor Wager's collaborators combined twenty experimental fMRI studies. It found reliable placebo-related changes in networks associated with cognitive and affective processing, with smaller and less consistent effects in some regions closely tied to nociceptive processing. The result argues against a simple story in which expectation uniformly turns down a single pain center.
Imaging limitations matter. BOLD signals are indirect measures of neural activity. Region names are not unique functions. If participants report less pain, changes in decision, memory, or reporting can accompany changes in sensation. Researchers use multivariate patterns, spinal imaging, pharmacological blockade, and behavioral controls to separate mechanisms, but no scan reads subjective pain without context.
Nocebo studies offer complementary evidence. Warning about side effects can increase symptom reporting, and negative treatment expectations can reduce analgesic benefit. Ethical communication should convey material risks without suggestive repetition or concealment. The goal is accurate informed consent, not manipulation.
Clinical benefits are generally modest and variable
Placebo-controlled trials estimate a treatment's effect beyond the comparison condition, not the fraction of improvement caused by expectation. Participants in both groups receive attention, assessment, and the possibility of recovery. A large change in a placebo arm can reflect natural history as well as contextual effects.
Cochrane reviews comparing placebo with no treatment found no large effects across diseases, while pain and nausea showed possible modest subjective benefits with considerable variation. Sham procedures can create stronger expectations than pills, but they also raise ethical and methodological issues. Evidence is weaker for objective disease outcomes.
Open-label placebo trials tell participants that pills contain no active drug while providing a rationale about learned responses. Some randomized studies in chronic low back pain, irritable bowel syndrome, and other conditions report symptom improvement over usual care. Meta-analyses find promising but heterogeneous effects, often in small studies with subjective outcomes, short follow-up, and researcher enthusiasm that may influence expectations. Larger preregistered trials and active communication controls are needed.
The ethical advantage is clear: open-label use avoids deception. The clinical case remains bounded. An inert pill should not replace antibiotics for bacterial infection, insulin for type 1 diabetes, surgery for an emergency, or evidence-based analgesia when indicated. Context can be used to support treatment through clear explanations, trust, and predictable routines without pretending that ritual cures pathology.
What is established and what remains open
Controlled experiments establish that expectations and learned cues can alter reported pain and, in some settings, physiological processing. Endogenous opioids and descending modulation contribute to part of this effect. Nocebo expectations can worsen symptoms. The size and mechanism depend on the person, procedure, and pain condition.
Contested questions include how much laboratory placebo analgesia transfers to persistent clinical pain, whether open-label effects survive stringent communication controls, and which biomarkers predict response. It is also unknown how to use context without increasing inequity or implying that patients are responsible for symptoms.
The evidence permits a practical conclusion that stops short of cure claims. Treatment context is one component of pain care and can augment or undermine an active therapy. It should be designed ethically and measured separately from natural improvement. Pain remains real when context changes it, just as vision remains real when lighting changes what is seen.
Sources
5 sources, none of them with a link, so treat them as reading, not proof.
- Levine, Gordon, and Fields, “The Mechanism of Placebo Analgesia,” The Lancet, 1978.
- Wager and collaborators, individual-participant meta-analysis of neural systems underlying placebo analgesia, Nature Communications, 2021.
- Colloca and Barsky, “Placebo and Nocebo Effects,” New England Journal of Medicine, 2020.
- Cochrane Collaboration, “Placebo Interventions for Clinical Conditions,” systematic review.
- Benedetti, Placebo Effects: Understanding the Mechanisms in Health and Disease, second edition, 2014.
