Upper Back Pain: Causes, Warning Signs and What the Evidence Says About Treatment
Table of Contents
Key Takeaways
- Upper back pain is common: in working adults, the one-year prevalence of thoracic spine pain ranged from 3.0% to 55.0%.[1]
- Disc herniations, bulges and annular tears are common on thoracic MRI even in people without thoracic symptoms, so imaging findings need to be read in context.[2][3]
- Not all upper back pain comes from the back: thoracic spine pain can also arise from the gastrointestinal, cardiopulmonary and renal systems.[3]
- In placebo-controlled research, spinal manipulation showed a small, clinically unimportant short-term advantage over placebo in one systematic review and no significant difference from sham (de-tuned ultrasound) treatment in a later trial, in which all groups improved over time.[4][5]
- In my clinical experience, upper back pain most often turns out to be muscular pain between the shoulder blades linked to computer work and static postures, but I also look for other explanations, such as soreness after coughing, shingles, a disc problem or pain referred from the stomach or the aorta.
Introduction: Upper Back Pain
Upper back pain receives less research attention than neck pain and low back pain. A systematic review of 33 studies pointed out that the thoracic spine has received less attention in clinical, genetic and epidemiological research, even though pain in this region can be equally disabling.[3]
It is not a rare problem. In a review of working adults, the one-year prevalence of thoracic spine pain ranged from 3.0% to 55.0%, and most occupational groups had medians of around 30%.[1]
This article looks at upper back pain from both a medical and a sports medicine perspective: what the term covers, where the pain can come from, which features suggest something other than a muscle or joint problem, and what randomised trials actually show about common treatments.
In my practice, upper back pain is a fairly heterogeneous group. When a patient tells me their upper back hurts, it can mean many different things: for some it means the spine, for others the neck or the shoulder region, and for some patients it can even mean the side of the chest. Most patients can tell me whether the pain is in the upper or the lower back, but sometimes a patient simply says “back pain”, which to a clinician’s ear sounds very broad, and then I often need to ask more specific questions.
What Counts as Upper Back Pain, and How Common Is It?
In research, upper back pain is often studied as thoracic spine pain. Briggs and colleagues defined it as pain felt between the first and twelfth thoracic vertebrae (T1–T12) across the back of the trunk.[3] In everyday language, people also use “upper back” for the area between the shoulder blades and the top of the shoulders, which overlaps with neck and shoulder pain.
Prevalence figures vary widely, and the review found that they depended on how thoracic spine pain was defined. When the definition was simply “any thoracic spine pain”, one-year prevalence in the general population ranged from 15.0% to 27.5%. Studies generally reported higher prevalence in children and adolescents, and particularly in females.[3]
Several factors have been linked to this type of pain. In the general population, thoracic spine pain was significantly associated with concurrent musculoskeletal pain and with growth, lifestyle, backpack, postural, psychological and environmental factors.[3] In working adults, associated factors included high workload, high work intensity, perceived ergonomic problems at work and high mental pressure.[1]
Mobile phone use has also been studied. In a systematic review of 18 studies on mobile phone users, neck and upper back complaints had the highest prevalence rates, ranging from 55.8% to 89.9%.[6] The authors noted, however, that the evidence is limited because it comes from cross-sectional and case-control studies.[6] Such designs can show associations but not cause and effect.
In my clinical experience, most patients with upper back pain mainly have sore muscles. Usually this is pain between the shoulder blades, which I see mostly as muscular pain, typically related to computer work and static work postures. It often goes together with tightness in the neck and shoulder region, and many of these patients also have headaches. The pain can be one-sided in many patients, depending somewhat on the type of work they do, and I have noticed that in right-handed patients the tightness is more often on the right side.
Where Does Upper Back Pain Come From?
Upper back pain does not always start in the spine. According to Briggs and colleagues, thoracic spine pain can arise from several sources, including thoracic and cervical spinal structures and the gastrointestinal, cardiopulmonary and renal systems. The thoracic spine is also a common site for inflammatory, degenerative, metabolic, infective and neoplastic conditions.[3]
Some of the causes I see are outside the spine. A patient may have the flu, for example, and the coughing can make the upper back sore through the intercostal muscles; in my experience, this is a fairly common cause of upper back pain. From a clinical perspective, intercostal nerve entrapment is another possibility.
Even so, a structural finding does not always explain the pain. The same review notes that, as in the lumbar spine, degenerative changes seen on thoracic spine imaging are not necessarily associated with pain.[3]
An MRI study of 90 people without thoracic symptoms shows why. Sixty-six of them (73%) had anatomical findings at one or more levels of the thoracic spine. These included disc herniation in 37%, disc bulging in 53%, an annular tear in 58% and deformation of the spinal cord in 29%.[2] Findings like these are common in people without thoracic symptoms, so imaging findings alone cannot establish the cause of an individual’s pain.
Then there are pain conditions arising from the spine itself, which I find diagnostically more difficult. These patients may have midline tenderness, and I usually image them once. The history matters: did the pain start with or without trauma, and is there reason to suspect a bone problem such as a compression fracture? I see compression fractures every now and then, but usually after trauma. In older patients I may see wedge vertebrae; in my experience there is often trauma behind these too, but sometimes previously unrecognised compression or wedge fractures show up on imaging. Occasionally a pathological process in the spine may also explain midline tenderness, and in these cases imaging becomes all the more important.
When real diagnostic imaging is needed, in my experience it is often MRI that gives more precise anatomical information, because the pain may also come from soft tissue, for example a disc. In my practice, the pain can fairly often be disc-related, although, as noted above, disc findings on MRI are also common in people without thoracic symptoms. Sometimes the pain radiates, and I consider a thoracic disc herniation pressing on a nerve when the pain is band-like and spreads towards the front of the chest.
Upper Back Pain Warning Signs: When It May Not Be Musculoskeletal
Serious causes are uncommon, but they matter. An international framework on red flags for serious spinal pathology, developed under the International Federation of Orthopaedic Manipulative Physical Therapists, states that serious spinal pathology is rare but can have devastating or life-limiting consequences and needs to be identified early. It also notes that high-quality evidence for the diagnostic accuracy of most individual red flags is lacking.[7]
Internal organs can also refer pain to the upper back. In one randomised trial on thoracic spine pain, people were excluded if they had an active history of visceral conditions that refer pain to the thoracic spine, listed as myocardial ischaemia, dissecting thoracic aortic aneurysm, peptic ulcer, acute cholecystitis, pancreatitis, renal colic and acute pyelonephritis. The same trial excluded people with osteoporosis, thoracic fracture, spinal infection, neoplastic disorders or spondyloarthropathy.[5]
Acute aortic dissection is among the most serious of these. In the International Registry of Acute Aortic Dissection (464 patients), sudden onset of severe, sharp pain was the single most common presenting complaint, while classic examination findings such as aortic regurgitation and pulse deficit were found in only 31.6% and 15.1% of patients. In-hospital mortality was 27.4%.[8]
In some patients, upper abdominal pain can also radiate to the back. In my experience, dyspepsia and heartburn in particular can be clearly felt on the back side. A clinical emergency always comes to my mind as well, because aortic dissection can also be felt in the back. When I am about to interpret upper back pain as purely muscular, I keep these visceral causes in mind and exclude them when needed.
I also often check the patient’s skin, because in my experience one-sided upper back pain surprisingly often turns out to be shingles. These patients usually have characteristic skin lesions, but not always. I have seen several cases where shingles was strongly suspected without any blisters, or was strongly suspected only afterwards. Confirming it is difficult: at least in our practice the sample is taken from a blister, so without blisters the diagnosis often remains unclear.
For me, upper back pain is somewhat different from low back pain in terms of how broadly I examine the patient. With upper back pain I usually also check the neurological status, at least of the upper limbs, especially if there is any neck pain, because sometimes the cause can be a cervical disc problem or another neurological condition affecting the arms.
Upper Back Pain in Athletes
In some sports, pain around the upper back and ribcage has a specific cause worth knowing about. Rowing is one well-documented example. In a cohort study of 151 Australian Rowing Team athletes covering the 2012–2016 (Rio) Olympiad, 19 rib stress injuries were identified (12 bone stress reactions and 7 fractures). The most common locations on imaging were the mid-axillary line and the sixth rib, and period prevalence ranged from 4% to 15.4%.[9]
These injuries were costly for training. A stress fracture led to a median of 69 days, and a bone stress reaction to a median of 57 days, lost from full on-water training. MRI was the most commonly used imaging method for diagnosis.[9]
The general-population data also matter for younger athletes: participation in specific sports was among the factors associated with thoracic spine pain in children and adolescents.[3]
What the Evidence Says About Upper Back Pain Treatment
The honest summary is that research on its treatment is thin. A systematic review by the Ontario Protocol for Traffic Injury Management (OPTIMa) collaboration concluded that quality evidence on managing musculoskeletal thoracic pain is sparse.[4]
Spinal manipulation. In the same review, spinal manipulation compared with placebo was associated with a small, short-term reduction in pain intensity that was not clinically important, and acupuncture produced outcomes similar to placebo. For recent-onset chest wall pain, a multimodal programme (manual therapy, soft tissue therapy, exercise, heat or ice and advice) produced a statistically significant but clinically unimportant short-term pain benefit compared with a single education session, and participants receiving multimodal care were more likely to report important improvements in chest pain.[4]
A later placebo-controlled trial supports this picture. Crothers and colleagues randomised 143 people with non-specific thoracic spine pain to spinal manipulative therapy (36), Graston instrument-assisted soft tissue therapy (63) or de-tuned ultrasound as a placebo (44), with up to 10 sessions over four weeks. There were no significant between-group differences in pain or disability at 1 week, 1 month, 3 months, 6 months or 12 months, but all groups improved over time.[5]
When manipulation was added to exercise in an open-label trial of 100 patients, the manipulation group had better pain scores after the eighth session (mean difference 1.14), but there was no significant difference in pain or quality of life at the 12-week follow-up.[10]
Exercise and strength training. The exercise evidence summarised here comes from the neck and shoulder region rather than the thoracic spine itself. In a randomised trial of 447 office workers, supervised high-intensity strength training for the neck and shoulder muscles was given as 1 × 60, 3 × 20 or 9 × 7 minutes per week for 20 weeks. Among participants with a baseline pain rating of 3 or more (n = 256), all three training groups reduced neck pain significantly compared with the reference group, and the authors concluded that one weekly hour of specific strength training effectively reduced neck and shoulder pain in office workers.[11]
Because many people describe pain between the shoulder blades and at the base of the neck as “upper back pain”, this research is relevant, but it should not be read as direct proof for pain in the mid-thoracic spine.
Patients with basic muscular pain linked to static, office-type work I often refer to an occupational health physiotherapist. In my experience, isometric exercises often seem to suit these patients best. They may of course also use other physical therapy, massage and stretching, or visit a chiropractor, but I usually tell them I see these passive treatments as a complement to active physiotherapy rather than a replacement for it.
In my practice, anti-inflammatory painkillers, and sometimes muscle relaxants, are often prescribed for muscular pain, and many of my patients feel they benefit especially from muscle relaxants. Because muscle relaxants can cause drowsiness, I am used to suggesting that patients take them in the evening; during the day they may prevent driving, which often limits their use. At night many patients find the drowsiness may even make sleeping easier. I do not prescribe them for regular use, and my patients usually take them as short courses. The choice of medicine, its timing and duration, and any driving precautions always depend on the specific medicine and the patient’s own situation.
When patients start active physiotherapy, I usually emphasise that pain relief does not necessarily come right away. In many patients the pain gets somewhat worse at first, and I usually explain this in advance, because in my experience active rehabilitation is, in the long run, what properly helps many patients, especially when the situation is more chronic.
Upper Back Pain: Practical Points From the Research
- Imaging findings need context. Disc herniations, bulges and annular tears are common on thoracic MRI in people with no symptoms.[2]
- Manipulation showed small or no differences compared with placebo. Placebo-controlled thoracic research found small, clinically unimportant or no between-group effects for manipulation, and in one trial all groups improved over time.[4][5]
- Strength training has evidence in the neighbouring region. Workplace strength training reduced neck and shoulder pain in office workers.[11]
- Some pain is not from the back. Thoracic spine pain can also arise from the gastrointestinal, cardiopulmonary and renal systems, and sudden onset of severe, sharp pain was the most common presenting complaint in a large aortic dissection registry.[3][8]
- Athletes have sport-specific causes. In elite rowers, rib stress injuries are a documented problem and can mean weeks to months away from full on-water training.[9]
Our occupational health physiotherapists also often visit the patient’s workplace to look at work ergonomics. In my experience, upper back pain, especially static muscle-tension pain, may already ease with this step alone in some patients.
Conclusion
Upper back pain is common, with one-year prevalence in working populations ranging from 3.0% to 55.0%,[1] yet it has received less research attention than other spinal regions.[3] Abnormal-looking MRI findings are common even in people without pain, and degenerative imaging findings are not necessarily associated with pain.[2][3]
The treatment evidence is modest. Spinal manipulation, instrument-assisted soft tissue therapy and acupuncture have shown small or no advantages over placebo in the available thoracic spine trials, and in one trial participants in all groups improved over time.[4][5] Strength training has trial support in the neck and shoulder region.[11]
In my view, the most important practical skill is recognising when upper back pain is not a back problem at all. Understanding that distinction, and knowing that participants with non-specific thoracic spine pain in one placebo-controlled trial improved over time regardless of treatment group, may help patients and athletes set realistic expectations.
This matches what I see in practice: most of my patients with upper back pain have muscular pain related to static work, and in my experience active rehabilitation is what helps many of them in the long run, even if the pain gets somewhat worse at first. Some patients, however, turn out to have something else, such as shingles, a spinal problem or referred pain from internal organs, which is why I keep the examination broad.
References
- https://doi.org/10.1539/joh.k8007
- https://doi.org/10.2106/00004623-199511000-00001
- https://doi.org/10.1186/1471-2474-10-77
- https://doi.org/10.1016/j.jmpt.2015.06.001
- https://doi.org/10.1186/s12998-016-0096-9
- https://doi.org/10.1016/j.msksp.2020.102196
- https://doi.org/10.2519/jospt.2020.9971
- https://doi.org/10.1001/jama.283.7.897
- https://doi.org/10.1136/bjsports-2019-101584
- https://doi.org/10.1155/2023/7537335
- https://doi.org/10.1136/bjsports-2011-090813
