Detailed guides to painful problems, treatments & more

Can the Mind Freeze Shoulders? Five Studies (Member Post)

 •  • by Paul Ingraham
Get posts in your inbox:
Weekly nuggets of pain science news and insight, usually 100-300 words, with the occasional longer post. The blog is the “director’s commentary” on the core content of a library of major articles and books about common painful problems and popular treatments. See the blog archives or updates for the whole site.

Luise Hollmann of the Sydney School of Medicine published one of the coolest papers I’ve cited in my frozen shoulder book, or anywhere else. In 2018, she and her co-authors reported some mind-blowing evidence that at least some frozen shoulder patients have way, way more shoulder mobility while anaesthetized.

That just isn’t possible if the joint capsule is truly stuck — mechanically seized up — the way we’ve always assumed it is. So maybe some frozen shoulders aren’t so “frozen” after all.

The paper is titled: “Does muscle guarding play a role in range of motion loss in patients with frozen shoulder?” My sensationalized version: “Has everyone been wrong about how frozen shoulder works? For decades? Seriously?!” If the shoulder can seize up spectacularly from muscle guarding… implications.

This is bigger than just frozen shoulder. There may be lessons here for the whole body, and many other conditions. (Hard not to think about vaginismus, just a week after Vaginismus Awareness Day. Talk about “guarding”…)

B&W photo of a woman’s shoulder, upper arm, and upper chest.

Muscle guarding may freeze shoulders. If so, we aren’t sure why the muscles do it. Is it a psychological thing?

There has not yet been any more research of this kind, alas, but today I will look at some new “circumstantial” evidence, five other studies that shed some light on the subject — the psychological side of the equation. If shoulders are frozen by muscle, does that mean we can think the stiffness away? Is that muscle-powered freezing sensitive to our psychology?


PainSci Member Login

Submit your email to unlock member content. If you can’t remember or access your registration email, please contact me. ~ Paul Ingraham, PainSci Publisher

Privacy & Security of this form This login is private and secure: the information you submit is encrypted, used only to search for matching accounts, and then discarded.

Why would muscles guard? What exactly do they think they are doing?

There are two flavours of muscle guarding, probably overlapping:

  1. More reflexive, lower-level neurology, possibly no brain involvement at all. This is analogous to being unable to hold your hand on a hot burner. Your nervous system just won’t allow it, regardless of your psychological state or intestinal fortitude.
  2. More psychological and emotional, higher-level neurology. Technically, movement is possible, just super hard, and only when unusually relaxed, comfortable, confident, and/or with great focus. This is more analogous to a rational fear of a more complex or unclear threat, like feeling paralyzed as you approach a drop-off, or trying to pick up a spider that is probably not dangerous.

This post is about the psychology-flavoured type of muscle guarding, because that’s the only type of evidence we have. I go where science shines a light. Even if it’s like a flashlight running low on batteries.

Muscle guarding doesn’t mean you’re a coward, afraid to move…

Muscle guarding probably does exist in some sense, but strictly speaking it has never been confirmed or explained. Even if it does happen, we don’t have any idea whether it is driven by psychological factors at all. There are many examples of unwanted and uncontrolled muscle contractions that are every bit as unyielding as a contractured joint capsule.

In fact, I would bet against muscle guarding being a by-product of psychological processes. I think it may be mostly or entirely out of our control for much the same reason that some chronic pain probably is: because the brain said so. If the brain decrees that movement is a bad idea, good luck changing its mind with yours! It really could be as futile as trying not to blink when you get dirt in your eye.

Also, and finally, even if it is “psychological,” that doesn’t mean we can do anything about it! A mentally healthy person can’t stick his hand in a blender either, and that’s not a spinal reflex — that sensible, potent inhibition comes entirely from the mind.

…but you could still have an "emotional relationship" with your shoulder issues!

Despite all that, psychology and emotion could be relevant to muscle guarding, and if they are related to how things go with frozen shoulder, that would be a valuable clue. And this is a thread that we can pull on, because we do have some research to help out with that.

The mind-pain relationship has been studied quite a bit over the years, generally finding that there is one (shocker, I know). I'll cover five papers here: two looking for the same signal in the shoulder, and three more in frozen shoulders specifically.

In 2018, Martinez-Calderon et al studied the links between beliefs about pain and their pain intensity and/or disability. What is a “pain belief” anyway? (You’ll be sick of that pair of words by the end of the post.) In this case, they studied catastrophization (fearing the worst) and kinesophobia (fear of movement), and expectations of recovery. Based on thin evidence, they “confirmed” the commonsensical: pain and pain beliefs tend to go together in the shoulder just as they do in other conditions.

They did not confirm which came first. And that’s important.

Shoulder physio goes a little better if you think it will

In 2019, De Baets et al entered the debate with the first of three papers covered in this post. De Baets seems to be quite interested in this question!

She began with a wide-ranging review of 21 studies, looking at how things go for people getting physiotherapy for their shoulder pain, identifying just two of several mental indicators that were prognostic. Expectations of recovery and resilience both predicted who will do better with physiotherapy. But not anxiety. Or depression. Or anxiety. Or coping skills, or somatization, or freaking out, or fear of activity. None of these things affected the outcome of regular medical care, and all of them were more of a factor with surgery patients.

Smoke from the muscle guarding fire? Perceived stiffness is not related to structure

In 2020,  Baets et al zoomed in on frozen shoulder with a small experiment with 17 patients, and this is now getting much more directly relevant to the muscle guarding question. They looked for relationships between several things: range of motion, perceived stiffness, pain, objective signs, and stuff going on between the ears.

Signs of disease were related to range of motion (which doesn’t seem surprising) but not to perceived stiffness — which seems much more interesting. Instead, how stiff people thought they were was related to pain … which in turn was related to both pain and mental state. So the headline there would be: actual mobility is related to the physical condition of the shoulder, but how frozen it feels is more about how much it hurts and what you think about it.

I think that’s one of the most interesting single points in all this mess.

There is a weak signal there indicating that psychology (“pain cognition”) might drive the feeling of stiffness independently of anything detectably wrong inside a shoulder. But it’s a weak, fuzzy signal — not nearly enough subjects, not enough measurements over time, and no power to identify causality. Just an inspiration to keep studying, mainly.

How frozen shoulders get linked to psychology

And De Baets did keep studying. In her next paper, later in 2020,  Baets et al declared a kind of “victory” with the decisive title “Pain-related beliefs are associated with arm function in persons with frozen shoulder.” I suspect a bias here (which is fine, researchers are allowed to believe things, if they test them properly). But they were fishing for a sign of psychology-powered muscle guarding… and they caught one.

This time, the experiment was even more focused, looking only at pain-related beliefs and arm function in 85 people with frozen shoulders. So we are really getting down to it now. If some shoulder freezing is caused by muscle-guarding — which is not a given — then this could show it.

Three kinds of “beliefs” were measured: fear, catastrophizing, and helplessness. They measured how variation in function was “explained” by non-psychological factors (36%), and then added in the mental stuff, which almost doubled the number to 62%.

Hence that conclusion-title.

Like all studies, this one has some significant limitations — it’s tricky just to know if you’re actually studying frozen shoulders and not some other kind of painful shoulder — but it is big and clever enough to be persuasive that psychology and shoulder freezing have some kind of relationship. And the authors believe that the psychology comes first (despite acknowledging that it is “not possible to infer a causal relationship” from this data).

And that’s the conclusion that an awful lot of healthcare professionals are going to run with. They will, as advised, consider the role of the mind “in order to optimize the medical, psychological and physiotherapeutic management of frozen shoulder.” They will concern themselves with their patients’ mental state.

So I hope De Baets et al. are right about the causality.

Not so fast: it could all be the other way around

There’s one last study, a very fresh Japanese experiment, that points the arrow of causality the other way, partially. This was a study of pain itself, not mobility and function, but relevant enough to take seriously.

Hirata et al chose an approach that was at least partially capable of identifying causality (finally). They found that pain made people feel helpless and fearful of the worst — imagine that! And not the other way around (pain and fear causing pain). With lots of technical disclaimers.

Ironically, these authors — just like De Baets et al. — clearly believe that it does work the other way around. They figure it works indirectly via other measurable psychological factors like fear and depression… but factors they didn’t measure in this study. They were trying to confirm a direct relationship, but it simply wasn’t in their data.

Using fancy stats on 93 patients, “this study was not able to reveal factors that directly affect pain intensity.” That almost directly contradicts their own expectations, and to some degree the conclusions of De Baets. So they reverted to “ah, must be indirect then” … but that can only be confirmed with another even better study.

What to make of all this

Here is the paper trail we’ve been on:

  1. Hollmann et al: Immobile shoulders were came unstuck under anaesthesia!
  2. Martinez-Calderon et al: Shoulder pain and negative pain beliefs tend to go together (in some way).
  3. De Baets et al: Identified a couple of minor signs that pain beliefs predict the outcome for physiotherapy for shoulder pain.
  4.  Baets et al: How stiff a shoulder feels is not related to structural factors, but to pain intensity and beliefs. Interesting!
  5.  Baets et al: Pain beliefs are linked to poor shoulder mobility.
  6. Hirata et al: Pain might actually cause catastrophization and helplessness and not the other way around.

That last one is really quite inconvenient for Team Mind-Freezes-Shoulder.

Not one stitch of this can confirm that muscle guarding causes some freezing, or that psychology in turn is what’s driving muscle guarding. Some of this evidence is somewhat consistent with muscle guarding that is sensitive to mental state, but you could equally say that it’s consistent with muscle guarding that is only a little sensitive to mental state.

Clearly we can only speculate without better data. This is where I put my nickel down on this topic for now:

  • Muscle guarding is contributing to at least some cases of frozen shoulder.
  • Pain and pain beliefs both cause each other modestly, but there are probably much stronger drives of shoulder pain than what you believe about it (like inflammation).
  • And mobility of the shoulder specifically — muscle guarding — is only slightly sensitive to psychology, if at all. It’s more likely to be a reflex-like behaviour.