Despite more than 100 years of documentation of a frozen shoulder (also called adhesive capsulitis), this condition remains as enigmatic as the “trigger point” phenomenon and the question “What is consciousness?” Typically, a frozen shoulder is where the shoulder joint feels stiff and painful and you could barely move your shoulder within its normal range of motion, like putting on a jacket. It sometimes flares up without warning and gradually worsens over time, but the symptoms fade away within one to three years—and sometimes longer to a decade or more. One day you are going about your business, and in one moment, there is a gradual or a sudden build-up of achy or sharp pain in your shoulder.
Unlike most types of shoulder pain, frozen shoulders have few obvious symptoms that allow physicians and physical therapists to pinpoint a diagnosis. Even with the lack of understanding, there is still some progress in pain research that can help steer clinicians and patients with shoulder pain to find the best treatment possible without wasting their time on less effective or ineffective treatments.
Frozen shoulders tend to occur among people who are over the age of 40, much more uncommon among teenagers and young adults. Women tend to be more likely to get frozen shoulders than men. While the scientific literature estimates that about 2% to 5% of most developed nations’ population has a frozen shoulder, the numbers can vary among more specific populations. For example, about 8% to 10% of the population in the U.K. has a frozen shoulder. In a sample population of more than 800 Asian Americans (mostly of Japanese descent), about 15.6% of this group has it. In Japan and China, it is commonly called a “50-year-old shoulder.”
Frozen shoulder anatomy
A quick review of the shoulder anatomy can give us the common understanding when symptoms of a frozen shoulder are discussed. The shoulder is made up the glenohumeral joint (the ball-and-socket joint), the acromioclavicular joint (acromion attaches to the clavicle or collarbone), and the sternoclavicular joint (clavicle attaches to the sternum). The head of the upper arm bone, or humerus, attaches to the glenoid fossa, a shallow cavity of the shoulder blade. Strong fibrous connective tissues keep the shoulder in place while allowing it to move freely in all directions.
Shoulder muscles that move the shoulder joint include the rotator cuffs (supraspinatus, infraspinatus, teres minor, subscapularis), teres major, deltoids, trapezius, rhomboids, pectoralis major and minor, upper arm muscles, and neck muscles (e.g. sternocleidomastoid, levator scapulae). Various ligaments and tendons attach the bones to the muscles or to other bones.
But the shoulders cannot without the help of the nervous system. Nerves that innervate the shoulders originate from the brachial plexus, a bundle of nerves that stem from the C5 to T1 of the spine. These nerves go through between the front and mid scalene muscles that are in the back of the neck, and branch off to different shoulder muscles.
In frozen shoulders, the anatomical focus would be primarily in the shoulder capsule and the glenohumeral joint as well as examining pain on a broader scale.
Frozen shoulder symptoms
The most common symptom of a frozen shoulder is the gradual onset of the “freezing,” “frozen,” and “thawing” stages. Contracture of the shoulder capsule—the gradual hardening or shortening of the connective tissues within and around the structure—is another common symptom. The duration of each stage varies among individuals, and it is difficult to predict how long it lasts.
Frozen shoulder stages
There are three common stages of a frozen shoulder: freezing, frozen, and thawing.
Freezing and painful
The “freezing” stage is likely the most excruciating stage of a frozen shoulder because it is likely a novel experience for many people and this may trigger a sense of danger in the shoulder. Shoulder pain often tags along with the freezing process, and it becomes a merry-go-round of pain, stress, and reduced range of motion. Although the scientific literature indicates that the freezing stage usually lasts between three to nine months, for those who are suffering, this duration can feel like an eternity—especially with the potential loss of the ability to perform most daily tasks, like reaching up to grab something in the closet or putting on clothes or a bra.
Frozen and stiff
The “frozen” stage is where there is a limit on how much shoulder movement you have in your shoulder, and pain usually is much less than the freezing stage. One primary symptom of the frozen stage is where the shoulder is protracted and internally rotated. Because frozen shoulders almost always occur on one shoulder instead of two, one shoulder would appear to be “slouched forward” if you are looking at yourself in the mirror. This phase lasts between nine to 18 months.
Thawing and recovery
The “thawing” stage is almost painless for many people and there is a gradual gain in the shoulder’s range of motion. However, this is the longest and the most varied stage of all, ranging from six months to nearly four years. In rare cases, it can last more than 10 years.
An early study in 1975 found that about half of the 49 patients studied never had full recovery of their shoulder range of motion. Only three patients had a life-long handicap of the shoulder. Interestingly, the researcher Brian Reeves found very little “defects” in the rotator cuff muscles among all of the patients regardless of the duration of the recovery.
A study in 2010 by Carbone et al. found that about 96% of the 85 patients with early stages of frozen shoulder or adhesive capsulitis had symptoms at the coracoid process, which is hook-like or bird beak-like structure (hence the name “coracoid”) at the lateral edge of the upper front part of the scapula. You could feel it by trace your finger from the middle of your collarbone to its edge toward the upper arm where the collarbone ends. That is approximately where most of the symptoms occur, at least in this study.
However, there is some overlap of the coracoid process symptoms with other shoulder problems in the study such as rotator cuff tears, tendon calcification, glenoid arthritis, and joint diseases at the acromioclavicular joint. These overlaps are quite small and the population sample varies (e.g. 485 rotator cuff tears vs. 85 adhesive capsulitis vs. 46 tendon calcification).
Frozen shoulder movement
Both active and passive range of motion is lost among frozen shoulder patients, and pain tends to worsen when the shoulder capsule is stretched, especially during the freezing stage. Overall, about 80% of all range of motion of the shoulder joint is lost, particularly in shoulder extension (when you put on a jacket) and horizontal adduction (lifting your arms away from your body like you are forming the letter T).
A 2003 study found various shoulder movement deficits among 10 middle-age subjects (nine women, one man) with frozen shoulders. The authors cautioned that the degree of internal rotation and external rotation are dependent on the shoulder position, such as adduction and abduction.
“The results raise questions about the validity of a theorized single capsular pattern of motion in these subjects,” they wrote.
Frozen shoulder causes
No one really knows for sure what is or are the main causes of frozen shoulders, but perhaps looking for the “cause” or “causes” may be the wrong approach. Like most types of pain, frozen shoulders are very likely caused by multiple factors rather than just one. Risk factors that may increase the likelihood of getting a frozen shoulder include diabetes, obesity, cardiovascular disease, metabolic syndrome, and various types of neurodegeneration. Hypothetically, these chronic conditions are associated with the dysfunction of the autonomic nervous system that may trigger various physiological conditions that lead to a frozen shoulder.
What is known about the “freezing” part is that there is a lack or slow down of turnovers of the extracellular matrix (ECM) of the joint capsule of the shoulder. The ECM consists of proteins and connective tissues that serve as “scaffolds” of a cell. It is also where cues are sent to cells and tissues to proliferate, differentiate, migrate, self-destruct, and a host of other functions. The ECM constantly undergoes remodeling, which is regulated by fibroblasts and a broad family of enzymes called metalloproteinases.
One kind of metalloproteinase is called matrix metalloproteinase (MMP), which breaks down collagen in the joint capsule, while another kind—called tissue inhibitor of metalloproteinase (TIMP) blocks MMP’s actions, as the name implies. Evidence in a 2013 Indonesian study found that two types of MMP levels were “significantly lower” and TIMP levels were higher among frozen shoulder patients than healthy patients. Thus, an imbalance of these enzymes may be a major contributor to a frozen shoulder. Many factors may instigate this process, but there has been hardly any strong evidence that indicates a causal relationship.
Inflammation of the synovial joint of the shoulder is another common symptom, which can contribute to pain if the shoulder is moved actively or passively. The thickening and contracture of the anterior shoulder capsule, particularly in the middle of the glenohumeral ligament and coracohumeral ligament, reduces the volume of the shoulder joint, which reduces shoulder movements and mobility.
Low-grade systemic inflammation and stress
There is some evidence that low-grade inflammation that hums along with you for many years can be a strong predictor of getting a frozen shoulder, but no causality of this relationship has been established. Physiotherapist and researcher Max Pietrzak from the University of Bath stated that such chronic inflammation is “strongly associated with upregulation” of several types of cytokines (broad category of proteins that modulate cell communication and immune function) among patients with a frozen shoulder.
According to Dr. Nicolas Rohleder, who is an Associate Professor of Psychology at Brandeis University, chronic inflammation can develop from acute stress that somehow becomes chronic. It is unknown why some people can recover from acute stress while others do not, but it is understood that depression, child abuse, work burnout and work-related stress, and other such factors determine how well or poorly some people respond to acute stress.
“Individual differences in one-time stress exposure are only one potential pathway from stress to disease,” Rholeder wrote. “In real life, humans are exposed to repeated stressful events, some of different and varying nature and others recurring and largely similar. It has been proposed that an adaptive way to respond to repeated exposures to the same stressful stimuli is to habituate, that is, to show lower psychosocial and biological responses to re-exposure.”
Pietrzak described metabolic syndrome as having too much abdominal fat tissue, having higher-than-normal levels of triglycerides, high blood pressure and fasting blood glucose, and low levels of the “good” cholesterol. He speculated that because frozen shoulders occur more frequently among older adults, metabolic syndrome and chronic inflammation may have the same underlying causes.
One study by Gumina et al. found that a majority of their 56 sample subjects had most of these metabolic conditions during the freezing stage. About 23% had high blood glucose, 64% had high “bad” cholesterol, 43% had high triglycerides, and 68% had high total blood cholesterol.
Diabetic frozen shoulder
Diabetics have at least a 10% to 30% chance of having a frozen shoulder, compared to 2% to 5% of the general population. The higher the number of hemoglobin A1c, the higher the chances of getting a frozen shoulder. One retrospective study by Chan et al. reviewed more than 24,000 patients with this type of hemoglobin and found that for every hemoglobin A1c level that is more than 7, there is an increased risk of 2.77% of getting a frozen shoulder.
Some research shows that pathology of the thyroid gland has a fairly strong association with frozen shoulders, particularly hypothyroidism. A 2020 Brazilian study compared the thyroid health and many health factors of 166 patients with a frozen shoulder with 126 patients who were diagnosed with rotator cuff tears and a control group of 251 subjects with no existing shoulder pain or problems. The researchers found that hypothyroidism and “benign nodules” on the gland are strongly associated with frozen shoulders.
About 17% of the patients in the frozen shoulder group had hypothyroidism and 10% had benign nodules, while the other two groups had 4% and 2% in the rotator cuff tears group and 8% and 2% in the control group.
While this finding does not mean having these conditions cause frozen shoulders, the study’s results are similar to previous research in Shanghai (China), Jerusalem (Israel), and Antalya (Turkey) on various thyroid problems with frozen shoulders. It is likely a combination of various physiological factors that lead to frozen shoulder syndrome.
While hypothyroidism is more common in the general population than hyperthyroidism, there is some evidence that it may also be another strong predictor of getting a frozen shoulder. A Taiwanese study of 4,472 patients in Taipei found that 162 of them had a frozen shoulder. When compared to other diseases, such as diabetes, hyperlipidemia, and hypertension, they found that those with hyperthyroidism had 1.22 times the risk of getting a frozen shoulder compared to the general population of Taiwan.
A few studies have shown some associations between frozen shoulders and stroke, but it is known that stroke and similar conditions can contribute to various kinds of shoulder pain. About 16% to 72% of stroke patients get shoulder pain on one side, and up to 80% of such patients have no voluntary control of their affected shoulder.
A 2013 study of 106 stroke patients in Nanjing, China, found that frozen shoulders (and complex regional pain syndrome) tend to start in one to three months post-stroke, while centralized pain begins after three months.
But having a stroke does not necessarily mean you would likely get a frozen shoulder. Sometimes it could happen the other way around. A Taiwanese study of 575 frozen shoulder patients compared this group with 1,201 control patients. The researchers found that the frozen shoulder group had nearly 1.5 times the likelihood of getting a stroke than the control group. Psychosocial factors, such low socioeconomic status, living in rural areas, being female, sleep deprivation, and sedentary lifestyle can also contribute to development of a frozen shoulder.
Parkinson’s disease is a neurological disorder in the brain that causes your body (particularly the limbs) to shake, which makes performing gross and fine motor movements difficult, including walking, stair climbing, and eating with a utensil. This is caused by the degeneration of neurons in the brain that produces dopamine, a neurotransmitter that is necessary for generating movement and coordination.
Papalia et al. in 2018 hypothesized that there may be a “causative loop” between shoulder stiffness and postural change of the shoulder that leads to inflammation and pain, which recycles back to the stiffness and protectiveness. This idea is partially based on the researchers’ clinical observations and experience where their patients who had shoulder pain and stiffness “were subsequently diagnosed with Parkinson’s disease.”
Because there is a lack of clinical studies, they cannot confirm the validity of their hypothesis.
There is not a lot of evidence that connects the relationship between frozen shoulders and cardiovascular disease. Researchers from the National Institute of Cardiology Ignacio Chávez in Tlalpan, Mexico, reported in 1994 that in a seven-year-period, five women and two men came to the hospital for persistent shoulder pain at the upper arm after they had a cardiac catheterization. They had limited shoulder movement, and four of them had diabetes. Considering that the hospital performs about 125 cardiac catheterizations in the upper arm per year, the researchers reported, such frozen shoulder incidences are “rare.”
Boyle-Walker et al. also found that heart disease and circulatory disease are more prevalent among 32 patients with frozen shoulders, along with diabetes and thyroid disease. The patients were compared with 31 people in the control group. However, they emphasize that their study “does not prove cause and effect” and
A 2000 study of 214 male patients who had a heart surgery at a Veteran’s hospital found that 35 patients reported shoulder pain and seven of them (3.3%) had adhesive capsulitis.
Since chronic low-grade inflammation is one of the primary factors of getting a frozen shoulder, perhaps cardiovascular diseases, such as atherosclerosis, should be taken into consideration when deciding you are at risk or not.
Broken bones in the shoulder, shoulder dislocation, and other injuries to the joint may contribute to getting a frozen shoulder, but the evidence that supports this relationship is scarce. One German study published in 1982 described several types of shoulder fractures and injuries among patients with symptoms of a frozen shoulder. Like other factors, there is no established causal relationship between physical trauma and frozen shoulders, and the latter may develop independently from the injury, according to the authors.
A group of bacteria called Propionibacterium acnes may contribute to the development of frozen shoulder. A pilot of study published in 2014 found that eight out ten patients had bacteria in their joint capsule tissues. Six of these bacteria cultures were Propionibacterium acnes. The researchers failed to find a relationship between the bacteria and frozen shoulder.
Researchers Bunker et al. postulated that the bacteria could get into the shoulder capsule from the gum and teeth and into the bloodstream via the digestive system. “It is possible that frozen shoulder is more common in diabetics because of their reduced resistance to infection. And it is also possible that the reason frozen shoulder never recurs is a result of acquired immunity,” they wrote.
While these bacteria normally inhabit the sebum of the hair follicles just under the skin, it is possible that these bacteria can get into the shoulder capsule after getting a shoulder injection. However, some research fails to find such a relationship.
As mentioned in the Parkinson’s disease section, the posture of the shoulder that most people with Parkinson’s have may be part of the hypothetical cycle of the frozen shoulder. But for those without the disease, how much does posture matter in contributing to frozen shoulder?
Some may think kyphosis or rounded shoulders can cause a frozen shoulder or other types of shoulder pain, but the current evidence does not support this idea. A 2016 systematic review of ten studies found that “thoracic kyphosis may not be an important contributor to the development of shoulder pain,” but it is likely that it reduces the shoulders’ range of motion like the ability to scratch your upper back.
“Even if the studies had reported significant differences in thoracic kyphosis between groups, it would not have been possible to establish whether the thoracic hyperkyphosis preceded the shoulder symptoms or if the thoracic hyperkyphosis was a postural adaptation to shoulder pain,” Barrett et al. stated.
Given the overwhelming evidence mentioned in the scientific literature, frozen shoulder is more likely a physiological disorder rather than a biomechanical one.
Prognosis of frozen shoulder
The first—and probably the only—systematic review on the prognosis of frozen shoulder concluded that there is enough evidence for clinicians to reliably predict the course of a frozen shoulder based on physical examination and patients’ symptoms, based on seven qualified studies. Patients with chronic frozen shoulders who also have diabetes or frozen shoulders on both sides should consider surgery as an early intervention.
The researchers reported the shortest average duration of a frozen shoulder is 15 months and the longest time is 30 months. About 70% to 90% of patients usually “respond well” to non-surgical treatments.
The drawback with this review is that the bias level of the studies are inconsistent with different studies reporting different treatment protocols, lack of patients’ activity level documentation, and the length of the follow-up period.
Because the natural course of frozen shoulder has not been studied extensively, the researchers concluded that “it is difficult to carry out a high-quality systematic review of this issue.
Diagnosis of frozen shoulder
Physicians and physical therapists would likely use a differential diagnosis for frozen shoulder, which is a way to compare two or more conditions that may be mistaken for one another. For example, frozen shoulders share some symptoms as shoulder tendonitis, shoulder impingement, and other types of shoulder pain.
They would first review your medical history and—hopefully—allow you to share your experience with your shoulder. They would also do two types of shoulder range of motion tests: passive and active. As mentioned earlier, if you have pain during a passive range of motion test, where the clinician moves your arm for you, then it is likely a sign of a frozen shoulder.
One active range of motion test is the painful arc test, where you stand or sit and move your arms at your sides. Then you gradually raise your arms up to your sides like you are making a snow angel. Pain between 60 degrees and 120 degrees is usually a positive sign of a frozen shoulder.
Clinicians may also use imagings like an X-ray and ultrasound to see your shoulder joint for any signs of injury or disease and to make sure that you do not have other factors that can contribute to your shoulder pain.
Do not use this article as a substitute for your personal medical advice or make your own diagnosis. Consult with a qualified medical profession for the best course of treatment.
Best frozen shoulder treatment
Treatments for frozen shoulders should be individualized since everyone has a different medical history and narrative behind their pain. Evaluating the effectiveness and efficacy of certain treatments can help you decide the best course of action.
Do not substitute this article for your doctor’s advice and care or to self-treat. This is for information only based on current scientific evidence.
Corticosteroid injections are a common treatment for the early stage of frozen shoulder that provide short-term relief. While there are numerous trials and systematic reviews that examined this topic, the evidence indicates that the outcomes are mostly short term like oral medications.
One trial of 122 patients with a frozen shoulder were randomly picked to be in the following groups: injection to the shoulder joint capsule, combined joint capsule and interval injections, and sham injection. There was no difference in pain relief in the treatment groups. Pain reduction was significant at six weeks, and the relief lasted up to 12 weeks but not at 26 weeks.
Medications such as nonsteroidal anti-inflammatory drugs (NSAIDs) may be used to initially control the pain, especially if it disturbs sleep. There are no randomized controlled trials to examine the effects of NSAIDs for frozen shoulders and have no effect on the natural course of the condition.
Oral corticosteroids also have been shown to have short-term pain relief and allow a bit more shoulder movement, according to a 2006 systematic review of five trials with a total of 179 patients. Three of the trials, which are ranked high quality, do not find long-term benefits beyond six weeks.
Calcitonin is another medication, usually taken as a nasal spray, that may alleviate the inflammatory symptoms of frozen shoulders. Cho, Bae, and Kim reported that nasal calcitonin treatment had a “notable” improvement than the placebo group where both groups were given physical therapy treatment during the trial. However, they did not link or cite this randomized controlled trial.
Heat or cold treatments?
There is very little evidence for how well heat or cold therapies work for frozen shoulders. One 2008 study from Hong Kong found some benefit to heat therapy with stretching to the shoulder joint’s range of motion (except for shoulder flexion), but the heating group’s outcome is not significantly better than the stretching-only group.
For cold therapy, there appears to be not much difference between that with NSAIDs, injections, mobilization with physical therapy, and no treatment, according to a 1984 study. An earlier study in 1976 found that ice therapy could “shorten” the painful freezing stage of frozen shoulders, but it is no better than physical therapy alone.
Frozen shoulder exercises
A 2014 Cochrane Review found that exercise, when used with manual therapy, does not have better outcomes than corticosteroid injections in the short term for pain relief and better shoulder movement. “It is unclear whether a combination of manual therapy, exercise and electrotherapy is an effective adjunct to glucocorticoid injection or oral NSAID,” the authors wrote.
Even so, they found that manual therapy and exercise “may provide greater patient-reported treatment success and active range of motion.” But self-reported success does not necessarily mean that their frozen shoulder pain has improved.
Nutrition and diet
Because frozen shoulders are mostly a physiological issue and have ties with metabolic syndrome, it is likely that nutrition may play a role in managing the symptoms. However, there is no existing evidence that any kind of diet would work for reducing the symptoms of a frozen shoulder.
Frozen shoulder surgery and manipulations
Surgery and manual manipulation is usually recommended for frozen shoulders if conservative treatments do not bring any pain relief. It is up to you and your physician or physical therapist to determine the risks versus benefits of surgery.
Manipulation under anesthesia (MUA)
MUA is where a manual therapist stretches your shoulder and or moves it in a way that tears and loosens the fibrous tissues in the joint capsule while you are under general anesthesia. Although there is no established timeline for MUA treatment, Finnish researchers Vastamäki et al. suggested it for those with the condition between six to nine months after the onset of a frozen shoulder. However, current evidence suggests that MUA is no better than corticosteroid injections and hydrodilation “at best.”
Risks of MUA include bone fractures and tearing of the shoulder tendons and labrum.
Arthroscopic capsular release (ACR)
This type of surgery is the most commonly used for frozen shoulder surgical treatment, due to improvements of arthroscopic technology. It allows the surgeon to see what is going on in the shoulder joint capsule. Current studies suggest that ACR has positive long-term outcomes for both pain relief and shoulder movement. One Australian study in 2012 confirmed success of ACR among 43 patients during a seven-year follow-up. The affected shoulders were “comparable” to the healthy shoulders.
The primary risk for ACR is that there is a small chance of damage to the axillary nerve.
Frozen shoulder massage
Although there is a lack of evidence that shows the effects of massage therapy on frozen shoulders, massage may alleviate frozen shoulder symptoms in the short run. A South Korean systematic review assessed whether massage has any short-term or long-term effects for general shoulder pain. Among the 15 qualified studied in the review, researcher Yeon Young-Ran from Kangwon National University found that “effect size of short-term efficacy was large and robust, thereby supporting the hypothesis that massage is an effective treatment for reducing shoulder pain.” There is no evidence of long-term benefits.
However, the effect size of massage therapy is not significantly difference than other types of treatment, such as physical therapy, acupuncture, and rest, Yeon reported. She also pointed out limitations of this systematic review, such as small sample sizes, mixed types of massage, and short follow-ups.
Since touch itself has been shown to modulate pain, it is no surprise that massage therapy can alleviate the symptoms of frozen shoulders. Although massage cannot break the connective tissues that limit movement in the shoulder joint capsule (and it is also outside of most massage therapists’ scope of practice), it may reduce the anxiety and depression that often accompany a frozen shoulder.
Almost any type of massage would help alleviate shoulder pain as long as the therapist does not exacerbate your pain. Find what works for you and give feedback to your therapist about pressure and quality of touch. Even if massage therapy offers short-term relief, a few nights of uninterrupted sleep could mean a lot for those with persistent frozen shoulder pain.
A native of San Diego for nearly 40 years, Nick Ng is an editor of Massage & Fitness Magazine, an online publication for manual therapists and the public who want to explore the science behind touch, pain, and exercise, and how to apply that in their hands-on practice or daily lives.
An alumni from San Diego State University with a B.A. in Graphic Communications, Nick also completed his massage therapy training at International Professional School of Bodywork in San Diego in 2014.
When he is not writing or reading, you would likely find him weightlifting at the gym, salsa dancing, or exploring new areas to walk and eat around Southern California.