Basic Neurosciences · Paper A
CBT for chronic fatigue and pain
Cognitive behavioural therapy for chronic pain treats the response to the pain rather than the pain, which is why it changes disability and distress more than it changes pain; for ME/CFS, NICE NG206 (2021) rejects that model and allows CBT only as support for living with the illness. Once pain has outlived its injury, the thoughts around it ("if it hurts it is tearing") drive fear of movement, avoidance and deconditioning, and that loop is what CBTCognitive behavioural therapy, a talking therapy that works on the link between thoughts, feelings and actions.Read more → unpicks. The same tool then meets two different rulebooks in the United Kingdom: for chronic primary pain, NICE NG193 says consider CBT or acceptance and commitment therapy alongside exercise and advises against starting opioids; for myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS), NG206 withdrew graded exercise therapy, states that CBT does not assume abnormal illness beliefs and behaviours are the cause, and allows CBT only to help someone manage symptoms, never as a cure.
The cause of persisting disability here is behavioural, and it was mapped as a loop by Vlaeyen and Linton (2000, in the journal Pain), the fear-avoidance model. Pain that has continued for more than three months, without ongoing tissue damage to explain it, is appraised as a sign of harm (catastrophising). That appraisal produces fear of movement (kinesiophobia), fear produces avoidance and guarding, avoidance produces deconditioning, low mood and loss of role, and each of those lowers the threshold at which the next episode of pain is felt. A second pattern sits beside it: boom-and-bust, where a good day is spent overdoing things and the following days are lost to the flare, so the person's baseline drifts downward. Cognitive behavioural therapy attacks the loop at both ends. Cognitively, it tests the harm appraisal against evidence; behaviourally, it replaces avoidance and cycling with graded, paced activity, sleep scheduling and problem-solving. Two comorbidities keep the loop turning and belong in the formulation. Depression is common in chronic pain and shares its sleep and energy symptoms, and post-traumatic stress disorder and chronic pain maintain each other, a relationship set out by Sharp and Harvey (2001, Clinical Psychology Review): the pain acts as a reminder of the trauma, hyperarousal amplifies the pain, and avoidance serves both conditions at once. Drug treatment fits the picture at a specific point. Amitriptyline relieves pain at doses lower than those used for depression, by an action on descending pain modulation rather than on mood, and duloxetine is licensed for painful diabetic neuropathy; neither substitutes for the behavioural work.
A man in his forties sprained his ankle badly eighteen months ago. The scan was clean by month three. He still walks with a stick, will not go down stairs unaided, and has not been back to work. Ask him why and he says: if it hurts, it must be tearing. Nobody prescribed the stick. His fear did. That is the whole of this topic in one patient: the injury healed, the pain persisted, and the disability grew out of the response to the pain.
Cognitive behavioural therapy () for chronic pain is a structured psychological treatment that changes the thoughts and behaviours that turn a persistent symptom into disability, and it assumes the pain is real. In the United Kingdom the same therapy is delivered under two NICE guidelines with two different models behind it: NG193 (2021) for chronic primary pain, where CBT or acceptance and commitment therapy (ACT, a therapy that builds valued activity around the symptom rather than arguing with it) is something to consider alongside exercise; and NG206 (2021) for myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS), where CBT does not assume that abnormal illness beliefs and behaviours cause the illness, may only be offered to help someone manage symptoms, and is never presented as a cure.
The largest pooled evidence, the Cochrane review by Williams and colleagues (2020), combined 59 trials of for chronic pain in adults and found small effects on all three of pain intensity, disability and distress, with pain the smallest of the three. That is not a disappointing result; it is the result the model predicts. CBT was never aimed at the signal. If a stem promises that CBT will take the pain away, you can reject it; if it says CBT improves function and mood more than it relieves pain, it has described the evidence correctly.
Keep the ankle in front of you. In week one the sprain is real, the pain is an honest signal, and rest is the right answer. By month six the tissue has healed but the pain has not gone, and the man has learned a limp. Each step follows from the last in the fear-avoidance model of Vlaeyen and Linton (2000): a catastrophic appraisal of pain (catastrophising, the belief that pain means damage) produces fear of movement (kinesiophobia), fear produces avoidance and guarding, avoidance produces deconditioning, low mood and loss of role, and every one of those lowers the threshold for the next pain. A second pattern, boom-and-bust, runs alongside it: a good day is spent catching up, the next three days are lost, and the baseline drifts down.
works on both ends of the loop. The cognitive half tests the harm belief against evidence, for example the clean scan and the fact that walking yesterday did not cause damage. The behavioural half replaces avoidance and cycling with graded activity: a level that can be held on a bad day, raised slowly and predictably, plus sleep scheduling and problem-solving around work and relationships. What the picture cannot do is tell you about the drugs, and the exam likes one in particular. Amitriptyline relieves pain at 10 mg at night, titrated towards 75 mg, far below the antidepressant range, and the effect arrives within weeks without mood needing to change; duloxetine 60 mg daily is licensed for painful diabetic neuropathy. Neither replaces the behavioural work, and the descending modulation that explains them is covered on the pain pathways card.
Imagine a woman referred with widespread pain and exhaustion. Step one is primary or secondary: does a condition adequately explain the pain? Inflammatory arthritis or a neuropathy makes it secondary, and the condition is treated first. Nothing found, and pain for more than three months, makes it chronic primary pain under NG193, and the menu is a supervised group exercise programme, or ACT, a single course of acupuncture, and for adults an antidepressant (amitriptyline, citalopram, duloxetine, fluoxetine, paroxetine or sertraline); NG193 says do not initiate opioids, gabapentinoids, benzodiazepines, paracetamol or NSAIDs, and someone already on them is reviewed rather than stopped abruptly.
Step two is the question that switches rulebooks: does activity produce a delayed, disproportionate worsening of symptoms a day or so later that takes days to settle? That is post-exertional malaise (PEM), and NG206 requires it, alongside debilitating fatigue, unrefreshing sleep and cognitive difficulty, all present for three months, before ME/CFS is diagnosed. Once you are in NG206, graded exercise therapy (GET, a programme of fixed incremental physical activity) is not offered, the Lightning Process is not offered, energy management within the person's limits replaces pacing-to-improve, and is described as support for managing symptoms and distress.
NG206 is explicit about the model behind that. It says for ME/CFS does not assume that people have abnormal illness beliefs and behaviours as an underlying cause, and that no programme based on deconditioning or exercise-avoidance theories should be offered. The fear-avoidance loop therefore stays on the pain side of the line; it is not the account of fatigue.
Step three, in either rulebook, is to look for the two companions that keep the loop turning: depression, governed by NICE CG91 when it sits on a chronic physical condition, and post-traumatic stress disorder, which Sharp and Harvey (2001) showed maintains chronic pain and is maintained by it, so the trauma is treated alongside the pain.
Chronic pain is pain lasting more than three months. Chronic primary pain is the label when no other condition adequately explains it, and it includes fibromyalgia, chronic primary headache and complex regional pain syndrome. ME/CFS is suspected after six weeks of the four core symptoms in an adult and diagnosed at three months, with post-exertional malaise the feature you must ask for directly. For the loop itself, ask about beliefs (what does the pain mean?), avoidance (what have you stopped doing?), and cycling (what happens on a good day?).
For chronic primary pain, NG193: exercise, or ACT, one course of acupuncture, an antidepressant for adults, and no new opioid, gabapentinoid, benzodiazepine, paracetamol or NSAID. For ME/CFS, NG206: energy management, symptom-directed care, CBT only as support and never as a cure, and no graded exercise therapy or Lightning Process. For depression alongside either, CG91. For alongside either, trauma-focused therapy in parallel. For amitriptyline, the analgesic titration from 10 mg at night, with 75 mg the usual ceiling outside specialist care. Both guidelines reviewed on 6 September 2026.
Can you explain, without looking, which single symptom moves a patient from NG193 to NG206, what treatment that move removes, and why? The answer is post-exertional malaise, graded exercise therapy, and the fact that in ME/CFS activity is followed by genuine worsening rather than a fear of it.
treats the limp rather than the ankle, so it restores function more than it relieves pain, and NG206 differs from NG193 because it rejects the limp model for ME/CFS, where activity brings genuine harm rather than feared harm.