Why neck pain trials keep working on the mid back, and what the pain score misses

A trial that landed this week did something that reads, at first, like a category error. It went after chronic neck pain by working on the mid back, and then it went and measured the neck.

Fifty four adults with chronic non specific neck pain were randomised into two groups of twenty seven. One did cervical stabilisation exercise. The other did cervical stabilisation plus thoracic mobility work, six weeks of it, and at the end both groups were measured on pain, neck muscle strength, neck disability, posture, spinal mobility, and what the shoulder and arm could actually do (Journal of Back and Musculoskeletal Rehabilitation, 2026).

What it found, and what it did not

Both groups improved. That belongs first, because it is the least reported result in this whole field and the one people most need to hear.

Then the differences. Adding the mid back work produced no extra pain relief. None. It produced no extra neck muscle strength either. Had the trial stopped there, the headline would have been that thoracic mobility exercise is a waste of six weeks.

It didn't stop there. On four other measures the mid back group pulled clearly ahead: posture, how far the spine could move, neck related disability, and what the arm and shoulder could do. The widest gap of the four was in spinal mobility. The narrowest was neck disability, and even that one favoured the group that had barely worked on the neck. The pain number didn't move any further. What a person could do with themselves did.

The arc this sits in

Follow the line back and it has been drawing itself for years.

In 2019 a meta analysis pooled fourteen trials of thoracic spine manipulation for mechanical neck pain. Against standard care it favoured the thoracic work on both pain and disability in the short term. At long term follow up only one of those two survived, and it was disability (PLoS One, 2019). That same review also reported that thoracic manipulation was no better than a placebo version of itself, which is the sort of finding a field usually files quietly at the back and this one put in the conclusion.

In 2024 a second meta analysis came at the question from the other side. Six trials, comparing hands on treatment aimed at the neck against treatment aimed at the thorax or the junction between them. The answer came back flat. No difference in pain, no difference in disability, no difference in how far the neck could turn (Musculoskeletal Science and Practice, 2024). Certainty ran from moderate down to very low, so hold it loosely, but the direction is hard to miss.

Seven years, three different research designs, one shape. Working on the sore place is not reliably better than working somewhere else that connects to it. And whenever the pain number and the function number disagree, it is the function number that keeps holding on.

What I keep seeing

Ask someone with a stiff neck to turn and look behind them, then watch what stays still.

The ribs, usually. What I see again and again in classes is a person who can describe their neck in fine detail, which side, which movement, what time of day it is worst, and who has nothing at all to say about what happens between the shoulder blades, because that region stopped reporting in a long time ago. Turning becomes a neck job. The head twists on a torso that has quietly resigned from the movement, and the neck, asked to do the work of ten vertebrae with seven, lets you know about it.

This is why a Feldenkrais Method® lesson aimed at the neck so often spends most of its time nowhere near the neck. Awareness Through Movement® lessons work by making a quiet region noticeable again, in small range, slowly, with the effort taken out, so the parts that had opted out rejoin the pattern on their own rather than being stretched back into it. The lessons in the Feldy online movement program are built that way on purpose. Neck and shoulder tension rarely stays local for long, and the way a desk sets the head and the mid back against each other is the ordinary daily version of what this trial went to the trouble of measuring.

Fifty four people over six weeks is small, and six weeks is nowhere near long enough to know how any of it looks a year out. None of this is a reason to stop what your physiotherapist gave you. Gentle awareness work sits alongside clinical care rather than in place of it, and if someone is already tracking your neck, this is a conversation worth having with them. Yoga, Pilates and Tai Chi arrive at the same region by a different mechanism, and each has its own strengths.

What these trials keep pointing at, gently and repeatedly, is that a pain score is a narrow window onto a wide room.

So turn your head slowly to the left, only as far as it stays easy and comfortable, and stop there. Then notice whether your ribs came with you, or whether they stayed exactly where they were.