Chronic pain in later life: what has changed in how it is treated

Community surveys typically find persistent pain in 40 to 50 per cent of older adults, and higher rates in residential care. It is associated with reduced mobility, falls, sleep disturbance, depression, social withdrawal and loss of independence.

It is also under-reported. Older adults frequently do not raise it, on the reasoning that pain is expected at their age, that the doctor has more important things to address, or that the only offer will be tablets they do not want. Clinicians, in turn, often do not ask.

And in people with dementia, pain is dramatically under-recognised, because the person cannot report it. Agitation, resistance to care, reduced appetite and withdrawal are frequently the presentation, and are frequently treated with antipsychotics rather than analgesia — one of the more consequential errors in geriatric practice.

Why the pharmacology is harder

Every analgesic class carries more risk with age.

NSAIDs cause gastrointestinal bleeding, acute kidney injury, fluid retention and raised blood pressure, and interact with anticoagulants, ACE inhibitors and diuretics — the so-called triple whammy that precipitates renal failure. Most guidelines advise avoiding long-term oral NSAIDs in older adults. Topical NSAIDs, by contrast, have far lower systemic exposure and are a reasonable choice for localised musculoskeletal pain.

Paracetamol remains the usual first step on safety grounds, though its efficacy for chronic musculoskeletal pain is more modest than long assumed.

Opioids cause constipation, sedation, falls, cognitive impairment and respiratory depression, and tolerance develops. The evidence for long-term opioid therapy in chronic non-cancer pain is poor: the SPACE trial found no advantage over non-opioid analgesia at twelve months, with more side effects.

Adjuvants — gabapentinoids, tricyclics, duloxetine — have specific roles in neuropathic pain, and their own problems. Gabapentin and pregabalin cause dizziness, sedation and falls, need renal dose adjustment, and have been substantially over-prescribed for non-neuropathic pain where they do not work. Amitriptyline is anticholinergic and best avoided.

What non-drug treatment actually delivers

The evidence for non-pharmacological management in chronic pain is, for several modalities, better than for the drugs.

Exercise. The most consistently supported intervention across chronic musculoskeletal pain conditions. It improves pain and function, and the mechanism is only partly mechanical — regular activity modulates central pain processing. The main barrier is the entirely reasonable belief that movement will make it worse.

Cognitive behavioural approaches and acceptance-based therapy. Good evidence, particularly for reducing the disability and distress associated with pain rather than the intensity of the sensation. Pain management programmes combining education, exercise and psychological approaches outperform single-modality treatment.

Pain education itself. Explaining that chronic pain involves changes in the nervous system's processing — that hurt does not necessarily equal harm — measurably reduces disability in trials. It is also the intervention most likely to be omitted for lack of time.

Weight loss where relevant, particularly for knee and hip pain.

Sleep. Pain and insomnia reinforce each other in both directions, and treating the sleep problem often improves the pain more than adjusting the analgesia does.

The under-treatment problem

The move away from opioids was justified. What has sometimes followed is worse: patients left with untreated pain and nothing offered in place of the drug that was withdrawn.

Abrupt opioid discontinuation in long-term users causes withdrawal, distress and occasionally serious harm. Tapering should be slow, planned, agreed, and accompanied by something else — access to physiotherapy, a pain programme, adjuvant medication where appropriate.

Under-treatment is not a safer error than over-treatment; it is a different one, with its own costs in immobility, deconditioning, isolation and depression.

Pain in dementia

Worth stating separately because the failure here is systematic. Observational tools exist — PAINAD, Abbey Pain Scale — and are validated for use in people who cannot self-report. They assess breathing, vocalisation, facial expression, body language and consolability.

Where behaviour has changed in a person with dementia, a trial of regular simple analgesia is a reasonable and evidence-supported step. A cluster-randomised trial of stepwise pain treatment in nursing home residents with dementia found significant reductions in agitation. Treating the pain worked where treating the behaviour would not have.

What to ask for

A functional goal rather than a numerical one. Referral to physiotherapy or a pain management programme, which is available in more places than people realise. A review of whether current medication is achieving anything. And, if you are caring for someone who cannot tell you — consider that the agitation may be a symptom of pain rather than of the dementia.