Thoracic spine mobility: the reading we do not have

Thoracic spine mobility is the constraint most desk workers have and the one thing this battery cannot measure. That is an awkward sentence to publish on our own site, so it goes at the top rather than in a footnote.

The upper back is the segment a screen day stiffens hardest. It is also the segment Limber leaves blank. Both are true today, and the second one is our doing rather than yours.

The short answer

• The thoracic spine supplies extension and rotation, and it is where trunk rotation actually happens.
• It is the hidden limit on overhead reach. A flexed upper back holds the shoulder blade forward and the arm runs out of room early.
• Limber has no thoracic reading. The seated rotation test is designed and not built — in the app or on this site.
• The Spine group is therefore the one station of five that cannot be surveyed, and four of five is the honest ceiling of the whole battery.
• Nine of the catalogue's 56 movements work this segment. We work it hard and cannot read it at all.

Thoracic spine mobility, defined

Twelve vertebrae, from the base of the neck to the bottom of the rib cage. Every one carries a pair of ribs, and the ribs are why this section moves less than the neck above it and the lower back below it. A rib cage is a structure, and structures resist bending.

Rotation is the segment's specialty and its main claim on your attention. The lumbar spine turns very little, because the shape of its joint surfaces prevents it, so when you look behind you or reach across your body most of the turn comes from here. Extension is straightening out of the flexed shape a chair puts you in — the direction a desk day removes and the one overhead reach needs. Side bend comes coupled with rotation; you rarely get one without some of the other.

There is no single number to quote for thoracic rotation, which is the first honest problem here. Johnson and colleagues examined the reliability of thoracic rotation measurement in healthy adults in the Journal of Athletic Training in 2012. The reading depends on the position it is taken in — seated, half-kneeling, on all fours — and on whether the pelvis is genuinely held still. Two people measuring the same spine two ways get two answers.

Grade: moderate for the measurement when the protocol is fixed. Weak for any single normal value.

Why overhead reach ends here

Raising an arm to vertical takes roughly 180 degrees of travel, and the shoulder joint does not supply all of it.

The shoulder blade has to rotate upward and tilt backward across the rib cage so the socket stays underneath the arm. The blade sits on the ribs, the ribs attach to the thoracic spine, and a thoracic spine held in flexion holds the blade forward and down. The arm then meets its own limit early, and the lower back arches to make up the difference — which looks like overhead from the outside and is not.

This is also why chest and lat stretching so often changes nothing. The tissue being stretched was not the constraint. It was one segment behind, the argument chest opener stretches works through.

Grade: moderate for the mechanical role in overhead reach. Demonstrate it in ten seconds: slump, raise an arm, note where it stops; sit tall, raise it again, note the difference. What is thin is the step after that — the claim that training the segment changes anything beyond the segment.

The station we cannot survey

Here is the arithmetic, printed so you can check it rather than take our word.

The Range Score is the mean of five groups. One is Spine. Spine contains exactly one area, thoracic. That area is fed by exactly one test at full weight, the seated thoracic rotation. And that test does not exist.

Four of those five can be surveyed today. The fifth cannot be surveyed by anybody, on any surface we run. The app states it plainly — these are not available yet — and the arithmetic here does the same: an area with no test has no score, an unscored area never becomes a zero, and the composite is the mean of the groups that have a reading. Nothing is filled in.

So when a page here says the ceiling is four of five stations, this is the missing one. Not a rounding error at the edge of the battery — the segment the app's own desk work targets hardest, and the app cannot read it.

The test as designed

The design exists in full, which is why this is a build gap rather than a thinking gap. Stated as design, clearly labelled, because nobody can take it today.

The capture is a seated rotation with the pelvis fixed, turned to each side, read off the phone's gyroscope rather than judged by eye. The score anchors are 30° → 20, 45° → 45, 60° → 70 and 75° → 90, interpolated between and flat outside. It would be recorded per side, like the knee-to-wall, the shoulder reach and the hip flexor test.

None of that is available to you. It is printed so that when it is built, you can see it was not invented afterwards to fit a result. The wall angel, test seven, sits in the same position: designed, banded at 20 · 45 · 70 · 90, and not built either. Wall angels covers it as a movement, which is all it currently is.

The work, by name and dose

Nine of the 56 movements in Limber's catalogue are filed to the thoracic spine — as many as to the shoulders, and more than to any other area in the upper body. Whatever else is wrong here, the segment is not neglected on the training side.

The grammar the app applies to them: a dynamic movement is one repetition per breath cycle, six to twelve repetitions, inhaling four seconds and exhaling six. A hold runs 20 to 60 seconds at a mild stretch, never a sharp one. Dynamic work opens a session; the long calm hold closes it.

Rotation with the pelvis pinned — M043 seated, M039 lying — is the direction people fake most, by turning the whole trunk from the hips and calling the result an upper back.

What the evidence supports

The awkward parts first, because they are the parts that matter.

Nobody has shown that thoracic mobility work changes symptoms. There is no trial of the size or quality that would settle it. The mechanism is clear and the outcome evidence is not, and those are two different things.

The shape of your upper back is not a health finding. Christensen and Hartvigsen reviewed spinal curves against health outcomes in the Journal of Manipulative and Physiological Therapeutics in 2008 and found no association. Grob, Frauenfelder and Mannion asked the adjacent question about cervical curvature in the European Spine Journal in 2007 and found none either. Shapes vary widely among people with no complaints at all.

How stiff a back feels is not how stiff it is. Stanton and colleagues, in Scientific Reports in 2017, compared what people reported about back stiffness against what was measured mechanically, and the two did not match. Upper back tightness works through what follows from that.

The best-supported thing an aching desk-bound upper body can be given is not a stretch. Andersen and colleagues, in Pain in 2011, gave office workers with frequent neck and shoulder pain two minutes a day of progressive resistance training and measured a real reduction over ten weeks. No stretching programme in this region has a randomised trial with a clinical outcome to set beside it. It is printed here, on a page about mobility work, because it is true.

Grade: thin for thoracic mobility work as an intervention. Moderate for the mechanical role of the segment in overhead reach. Work it because you want the range, not because the range has been shown to buy anything else.

Pain that persists, wakes you at night, radiates into a limb, comes with numbness or weakness, or followed a fall or an injury is a matter for a clinician who can examine you. Nothing here diagnoses anything and nothing here is treatment.

Where to take a reading instead

You cannot measure this segment with Limber today. You can measure the chain it sits inside.

The shoulder mobility test is the Apley scratch, both sides, five bands: hands more than two hand-widths apart reads 10, one to two hand-widths reads 30, under one hand-width reads 55, fingertips touching reads 75, fingers overlapping reads 95. It sees the whole reach chain, including whatever the upper back contributed, and it cannot tell you which link was short.

The Range Score composes whatever you have taken into one 0–100 reading and draws the rest as unmeasured ground rather than guessing at it. Take every built test and it shows four stations surveyed and one blank. The blank is this one.

For the segment itself there is a home check that costs nothing and scores nothing. Sit on a hard chair, arms folded, turn each way without the pelvis following, and note where each side lands against a fixed object. Same chair, same method, six weeks apart. It is not a reading and this site will not dress it up as one.

The rest of the region sits under shoulders and upper back, and shoulder mobility exercises covers the joint at the top of the chain.

Questions

Does Limber measure thoracic rotation?

No. The seated thoracic rotation is test six of a seven-test design and it is not built, in the app or on this site. The Spine group is fed by one area, that area is fed by that one test, so Spine is the single station of five that cannot be surveyed. Four of five is the honest ceiling of the whole battery today.

How do I know if my upper back is stiff?

Not from how it feels — measured stiffness and reported stiffness did not match in the best experiment on the subject. A rough check is whether the segment will move when you ask it to: sit tall with the pelvis still and turn each way, then extend backward over a support. If the range appears on demand, the complaint is unlikely to be range.

How often should I work thoracic extension?

Often and briefly beats rarely and heavily, because the problem is usually time in one shape rather than a short structure. A daily minute or two is a sensible default, and the app's holds run 20 to 60 seconds. Treat that as reasoning from the mechanism, not a tested prescription — no trial has set a dose for this segment.

Will a foam roller change my upper back?

It gives you a support to extend over, which is the useful part. M044 is exactly that: lie back over a roller placed across the upper back, hold 45 seconds. What it does not do is remodel anything, and range gains from stretching reverse once the stretching stops.

Is a rounded upper back something to correct?

It is a shape, and shapes vary enormously in people who have no symptoms. A review of spinal curvature against health found no association, and the same result came back for the neck. What matters is whether the segment moves when you ask it to — a claim about range, which is testable, rather than about posture, which mostly is not.

Does thoracic mobility improve overhead reach?

Mechanically the link is clear: the shoulder blade sits on the ribs, the ribs attach to this segment, and a flexed upper back holds the blade where it limits the arm. You can feel it by raising an arm slumped and then sitting tall. What has not been shown is that training the segment produces a lasting change anywhere else. Treat the mechanism as sound and the outcome as unproven.