Neck Pain: Causes, Exercises and When Imaging Is Needed

The cervical spine consists of seven vertebrae, intervertebral discs, small facet joints, ligaments and a dozen or more layers of muscle that hold up the head (which weighs several kilograms) while allowing a very wide range of movement. Pain can arise from any of these structures, and often from several at once. In clinical practice this is described as 'mechanical pain': it worsens with certain movements or positions, eases in others, and the neurological examination is normal.

Why is there no obvious cause? First, muscle tension, overload of the small joints and hypersensitivity of the nervous system to pain signals do not produce an image that can be seen on an X-ray or MRI. Second, changes visible on imaging (degeneration, disc bulges) are very common in people who feel no pain at all — we discuss this in a separate section. Third, the course of pain is influenced by factors that do not show on a scan: sleep, stress, mood, activity level and worries about one's own health. The evidence for these factors is mostly observational, but its direction is consistent with what is seen in back pain at other levels of the spine.

Frequent, well-recognised clinical pictures include neck pain with an accompanying headache starting at the nape (so-called cervicogenic headache), pain with restricted head rotation after sleeping in an awkward position, pain radiating to the shoulder and arm, and pain combined with tingling or numbness of the hand. The last two require more careful assessment by a doctor because they may indicate involvement of a nerve root. As with other musculoskeletal complaints, it is important not to equate the presence of changes on imaging with the cause of the pain — we discuss this more broadly in our knowledge-base entry on magnetic resonance imaging.

The popular term 'tech neck' or 'text neck' describes neck pain attributed to looking down at a phone screen for long periods. The plausibility of this idea is partial. Biomechanically, tilting the head forward increases the load on neck structures — this is reported in biomechanical studies. Much less well documented is whether such a position by itself causes lasting pain and, if so, how strongly.

What follows from this? The result points to an association, not to a proven cause: retrospective studies cannot distinguish a situation in which the phone causes pain from one in which people with neck pain use the phone more (for example because of stress, poorer sleep or sedentary work). We classify this evidence as preliminary. A reasonable practical conclusion is that there are no good grounds to frighten people with the phone as a certain cause of disease, but there are reasons to limit long, uninterrupted periods in one position — whether it is a phone, a laptop or a steering wheel.

The situation with computer work is similar. Observational studies link long sedentary work with more frequent neck pain, but a single 'ideal' sitting posture has no strong support from randomised trials. Far better documented is the benefit of frequently changing position and taking breaks for movement. We write about the consequences of a sedentary lifestyle in a separate article; here we only note that ergonomic advice (screen at eye level, forearm support, regular breaks) is sensible and cheap, but its effectiveness in treating existing pain is only moderately supported.

Distinguishing these three situations is the most important diagnostic step, because it determines the pace and scope of investigation. Non-specific (mechanical) pain involves the neck and nape, may radiate to the shoulder or between the shoulder blades, but does not cause loss of strength or sensation and does not disturb gait.

Cervical radiculopathy is irritation of a nerve root, most often by a disc protrusion or narrowing of the intervertebral foramen related to degenerative changes. It produces pain radiating along the arm, often with tingling or numbness in particular fingers. Diagnostic reviews indicate that the accuracy of single physical tests (for example Spurling's test) is limited: in the systematic review by Thoomes et al. (Spine Journal, 2018) Spurling's test had high specificity (0.89–1.00), but sensitivity ranged from 0.38 to 0.97, and the authors judged that the evidence for the accuracy of physical examination tests is limited. In practice this means that the diagnosis rests on a combination of history, neurological examination and — if needed — imaging, not on a single test.

Cervical myelopathy (degenerative compression of the spinal cord) is less common but more dangerous, because its recognition is often delayed and spinal cord damage may be permanent. According to the review in the Journal of the American Board of Family Medicine (McCormick et al., 2020), reduced hand dexterity and gait instability, together with sensory and motor disturbances, are characteristic, and magnetic resonance imaging is the investigation of choice; the authors recommend prompt referral to a spine surgeon for any suspicion of this disease. For this reason, clumsy hands or an unsteady gait in a person with neck pain are treated as a signal for an urgent consultation, not for 'waiting it out'.

Exercise is the best-studied treatment for neck pain apart from drugs. The key evidence is the Cochrane update covering randomised trials in which exercise was compared with a control group in adults with neck pain.

The practical conclusion is twofold. First, in chronic pain it pays to train the muscles, not only to stretch them: strengthening the neck and shoulder-blade region has better support than stretching alone. Second, to be fair, for acute pain the review found no evidence — not because exercise does not work, but because there are no trials that would settle the question. In fresh, acute pain gentle movement within the limits of comfort is usually used, and more systematic training is introduced as the pain improves.

General strength training, which we discuss in our knowledge-base entry on strength training, can be a sensible addition, but the Cochrane review did not confirm that general fitness (for example walking) by itself changes neck pain in the short term. General movement remains recommended for health, sleep and mood, but it does not replace exercises aimed at the neck and shoulder blade.

Manual therapy is popular, and the evidence for its effectiveness is sometimes overstated. The Cochrane review on manipulation and mobilisation (Gross et al., 2015) included 51 trials (2920 participants). The authors found that results for manipulation and mobilisation of the cervical spine compared with a control group are few and diverse, whereas manipulation of the thoracic spine improved pain in acute and subacute neck pain (moderate-quality evidence; SMD -1.26 at short-term follow-up), although a funnel plot suggested possible publication bias. Multiple sessions of cervical manipulation might give better relief than some medications, and manipulation and mobilisation gave similar results.

The review authors also stress that there is a risk of rare but serious adverse events after neck manipulation and that better studies are needed. For this reason cervical manipulation (the forceful, 'cracking' techniques) should be carried out only by a qualified practitioner, after contraindications have been excluded, and not by an untrained person. Gentle mobilisation techniques and work on the muscles within physiotherapy are the safer option for most people.

The physiotherapy practice guideline for neck pain developed by the Orthopaedic Section of the American Physical Therapy Association (Blanpied et al., 2017) is based on the ICF classification and a review of the literature; this document illustrates a general principle well: treatment is chosen to match the clinical picture (with mobility deficits, with headache, with radiating pain) rather than applying one solution to everyone. It is worth remembering this when we hear that there is 'one method' for neck pain.

Painkillers (paracetamol, non-steroidal anti-inflammatory drugs) may briefly make movement easier but do not treat the cause, and their long-term use carries a risk of adverse effects; the choice of drug and duration should be decided by a doctor, especially in heart or kidney disease, ulcers or when taking anticoagulants. A soft orthopaedic collar has no good justification in non-specific pain when worn for long, because immobilisation promotes muscle weakening — advice on brief use applies to particular situations determined by a doctor.

The question about MRI is one of the most frequent. The answer begins with an inconvenient fact: abnormalities visible on MRI are very common in people with no symptoms, so the imaging result alone rarely settles the cause of pain.

This means that when someone with neck pain has an MRI, they will very probably see a 'bulging disc' — and the same finding is present in the vast majority of people without pain. Such a description can cause unnecessary anxiety, reinforce the belief in a 'damaged' spine and lead to avoidance of movement, which works against recovery. For this reason routine imaging in non-specific pain is regarded as unnecessary, and results must always be interpreted together with symptoms and a medical examination.

In other situations — that is, with typical neck pain after long sitting, after a night in a poor position or during a period of heightened stress — imaging usually does not change management. The exception is when the patient and doctor, after talking, agree that the result will genuinely help a decision. We write more about what MRI is, how the examination goes and what it cannot show in our knowledge base.

It is worth noting that the same set of principles (stay active, strengthen gradually, do not fear movement) appears in guidance on lower back pain; we discuss it in a separate article on exercise for chronic lower back pain. The neck and lower back differ anatomically, but the 'movement instead of rest' approach has a similar rationale.

It is also worth reporting to a doctor (without rushing, but not putting it off for months) when neck pain lasts longer than a few weeks despite movement and simple measures, when it is getting worse, when it wakes you at night or when it is accompanied by tingling in the hand. The doctor will assess whether referral to a physiotherapist or neurologist, or imaging, is needed.

Our analysis gives a consistent picture: neck pain is common, usually mild in course and responds well to an active approach, and the biggest risk is often not the condition itself but over-interpretation of an imaging result and avoidance of movement. At the same time, a small proportion of cases conceal more serious problems, which is why knowing the warning signs is as important as knowing the exercises.