Neck pain: what it is and what actually helps

Most neck pain that sends people looking for answers online is what clinicians call non-specific or mechanical neck pain: discomfort in the neck, often with restricted movement, that is not traced to a single structural cause like a fracture, infection, or tumor. A 2017 clinical practice guideline from the Orthopaedic Section of the American Physical Therapy Association groups most patients presenting with neck pain into four categories: neck pain with limited range of motion, neck pain related to whiplash, headache linked to neck pain, and neck pain with pain radiating into the arm from an irritated nerve root.^[1] This article focuses on the first and most common pattern, ordinary mechanical neck pain, while covering the arm-radiating and spinal cord warning signs that need a different response.

If your pain radiates into the leg instead, the relevant companion piece is our article on sciatica: what it is and what actually helps, which covers the equivalent nerve-root pattern lower in the spine.

Neck pain is one of the most frequent musculoskeletal complaints worldwide. An analysis of the Global Burden of Disease Study 2019 estimated an age-standardized prevalence rate of roughly 2,697 per 100,000 people, translating to about 222.7 million people living with neck pain globally in a single year, with rates highest between the ages of 45 and 74.^[4] It is common enough that most people will experience an episode at some point, which is part of why understanding what the evidence actually supports matters more than reacting to a single flare-up with alarm.

The clearest trial evidence for chronic neck pain concerns exercise. A Cochrane review pooling 27 randomized trials and 2,485 participants found moderate-quality evidence that cervico-scapulothoracic and upper-extremity strengthening produced moderate to large reductions in pain immediately after treatment and at short-term follow-up, while combinations of stretching and strengthening exercise for the neck, shoulder, and shoulder blade produced smaller but still favorable effects on pain and function over the longer term.^[2] Lower-intensity endurance training of the same muscle groups showed a smaller effect. The review authors were explicit that no high-quality evidence was found for any single approach, only moderate and lower grades of certainty, which is why exercise is described here as well-supported rather than proven beyond doubt.

The 2017 physical therapy clinical practice guideline builds on this same literature and recommends a combined approach: exercise paired with manual therapy (hands-on mobilization or manipulation performed by a trained clinician) and patient education, rather than any one treatment used alone.^[1] Passive treatments, meaning things done to a patient rather than active movement a patient performs, generally have weaker support when they are not paired with an exercise component.

As with low back pain, an early scan is often not the useful next step it seems like it should be. A study of 1,211 people with no neck symptoms at all found disc bulging on MRI in 87.6% of them overall, rising with age, along with measurable spinal cord compression in 5.3% and increased signal intensity, a marker sometimes associated with cord injury, in 2.3%, despite every participant reporting no symptoms.^[5] Even in people in their twenties, disc bulging was already present in the majority of scans.^[5]

This is why imaging is not usually the first step for ordinary mechanical neck pain: a scan frequently cannot distinguish an incidental, symptom-free finding from an actual source of pain. Imaging becomes more useful when arm-radiating symptoms suggest nerve root involvement, when red-flag symptoms covered below are present, after significant trauma, or when a specialist is weighing further treatment such as injections or surgery and needs a finding that would actually change the plan.^[3]

Manual therapy for the neck, including manipulation, has trial support for pain and function described above, but it also comes with a safety question that deserves an honest answer rather than a reassuring or alarming one. A systematic review of the available case-control studies concluded that conclusive evidence is lacking for a strong association between neck manipulation and stroke, but evidence is equally lacking to rule an association out.^[6] Some of the studies reviewed found elevated odds ratios, but the authors noted substantial bias and confounding, including the possibility that neck pain from an arterial dissection that had already begun led someone to seek manipulation in the first place, rather than the manipulation causing the dissection.^[6]

Serious vascular events after neck manipulation appear to be rare, but rare is not the same as ruled out. This uncertainty is one reason exercise and manual therapy combined, rather than manipulation used in isolation, has the stronger overall evidence base for chronic neck pain described earlier in this article.

Most neck pain, including pain that radiates into one arm, is not an emergency. A distinct pattern of symptoms, however, can indicate cervical myelopathy (compression of the spinal cord itself, rather than just a nerve root) or a fracture, and needs prompt assessment rather than a wait-and-see approach.^[3]

Pain radiating down one arm with tingling, on its own and without these features, more often reflects cervical radiculopathy (irritation of a single nerve root) and is usually managed without emergency intervention unless weakness is clearly progressing.^[3] When in doubt, a same-day call to a clinician is a reasonable middle ground between ignoring new neurological symptoms and treating every ache as an emergency.

Ordinary neck pain is common, usually mechanical rather than structurally damaging, and responds best to the same broad pattern seen across musculoskeletal pain conditions on this site: targeted exercise, paired with manual therapy and education where useful, ahead of routine imaging or passive treatment alone. The uncertainty around manipulation and stroke risk is worth knowing rather than dismissing, even though serious events appear rare. See how we grade the strength of evidence across topics in our editorial and evidence standards, and browse the rest of our Mobility, Posture & Pain coverage as it grows.

None of this replaces an individual assessment from a qualified clinician, particularly if any of the red-flag symptoms above apply to you, or if pain and function are not improving after a reasonable period of active self-management.